Plasma concentration and lipoprotein distribution of ApoC-I is dependent on ApoE genotype rather than the Hpa I ApoC-I promoter polymorphism

Transcription

1 Atherosclerosis 169 (2003) 63/70 Plasma concentration and lipoprotein distribution of ApoC-I is dependent on ApoE genotype rather than the Hpa I ApoC-I promoter polymorphism Jeffrey S. Cohn *, Michel Tremblay, Lucie Boulet, Hélène Jacques, Jean Davignon, Madeleine Roy, Lise Bernier Hyperlipidemia and Atherosclerosis Research Group, Clinical Research Institute of Montréal, 110 Pine Avenue West, Montréal, Québec, Canada H2W 1R7 Received 25 August 2002; received in revised form 14 February 2003; accepted 3 April 2003 Abstract An Hpa I restriction site located 317 bp upstream of the transcription initiation site of the apoc-i gene has been shown to increase apoc-i gene transcription in vitro. The aim of the present study was to determine whether this genetic polymorphism was associated in vivo with increased plasma levels of apoc-i. In a cohort of French /Canadians (n/391) recruited for a family study, we found strong linkage disequilibrium between the genes for apoc-i and apoe (as reported before for European/Americans), such that the apoc-i Hpa I-negative (H1) allele was strongly associated with apoe o3, whereas the apoc-i Hpa I-positive (H2) allele was strongly associated with apoe o2 and o4. ApoC-I and apoe were measured by ELISA in total plasma and in very lowdensity lipoproteins (VLDL) separated by ultracentrifugation (db/1.006 g/ml), and then by difference for the non-vldl fraction (d/1.006 g/ml), in a subset of families selected for their diverse apoe genotypes. Subjects were divided into normolipidemic (NL, n/89, TGB/2.3 mmol/l, LDL-CB/3.8 mmol/l) and hyperlipidemic groups (HL, n/88, TG/2.3 mmol/l and/or LDL-C/3.8 mmol/l). In NL subjects, apoc-i levels were not significantly associated with apoc-i genotype (H1/H1, H1/H2 or H2/H2). They were, however, related to apoe genotype, such that apoe3/2 subjects tended to have higher and apoe4/3 subjects tended to have lower concentrations of total plasma and non-vldl apoc-i and apoe. Total plasma, VLDL and non-vldl apoc- I and E levels were also higher in HL subjects with an apoe2/2 or apoe3/2 genotype. These results suggest that plasma levels of apoc-i are more strongly influenced by apoe genotype than by the Hpa I apoc-i promoter polymorphism, which probably reflects an effect of different apoe isoforms on plasma lipoprotein and plasma apoc-i metabolism, rather than a direct effect of apoe alleles on apoc-i transcription. # 2003 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Triglyceride; Cholesterol; Atherosclerosis; Lipoprotein metabolism 1. Introduction Apolipoprotein (apo) C-I is recognized as having a number of different lipid-regulating functions. It is involved in the regulation of plasma triglyceride (TG) Abbreviations: Apo, apolipoprotein; EDTA, ethylenediaminetetraacetate; HDL, high-density lipoprotein(s); LDL, low-density lipoprotein(s); TRL, triglyceride-rich lipoprotein(s); VLDL, very low-density lipoprotein(s). * Corresponding author. Tel.: / ; fax: / address: cohnj@ircm.qc.ca (J.S. Cohn). levels through inhibition of triglyceride-rich lipoprotein (TRL) binding and uptake by the low-density lipoprotein (LDL) receptor [1,2], the LDL receptor-related protein (LRP) [3,4], and/or the very low-density lipoprotein (VLDL) receptor [5]. It also reduces plasma cholesterol ester transfer protein (CETP) activity [6,7], and can inhibit adipose tissue uptake of free fatty acids [8]. ApoC-I has thus been implicated in the etiology of human hypertriglyceridemia, obesity and insulin resistance [8 /11]. The plasma concentration and lipoprotein distribution of apoc-i in humans has been investigated in a number of previous studies [12 /20]. After an overnight /03/$ - see front matter # 2003 Elsevier Science Ireland Ltd. All rights reserved. doi: /s (03)

2 64 J.S. Cohn et al. / Atherosclerosis 169 (2003) 63/70 fast, total plasma apoc-i concentration in healthy normolipidemic subjects is 4/12 mg/dl [12 /16], and is increased: (a) in hypertriglyceridemic but not hypercholesterolemic patients [13,15]; (b) in type III hyperlipoproteinemic patients having increased remnant lipoprotein levels [16], and (c) in diabetics [17]. ApoC-I levels are reduced in Tangier patients having significantly reduced levels of high-density lipoproteins (HDL) [18]. In the fasted state, about two-thirds of apoc-i in plasma is normally associated with HDL and one-third is associated with VLDL [13]. Plasma apoc-i concentration does not change significantly after the ingestion of a fat-rich meal, although the amount of apoc-i associated with TRL increases, while that in d/1.006 g/ ml lipoproteins decreases [19]. The postprandial increase in TRL apoc-i is largely due to apoc-i on smaller (Sf 20 /60) TRL containing apob-100 rather than TRL containing apob-48 [20]. A polymorphic Hpa I restriction site exists between the apoe and apoc-i genes on chromosome 19 [21]. It is located 317 base pairs (bp) upstream of the transcription initiation site of the apoc-i gene and is produced by a 4- bp CGTT insertion [22]. Xu et al. have shown with a reporter-gene assay that this 4-bp insertion (the H2 allele) is associated with a significant increase in apoc-i gene transcription. They proposed that the H2-allelic insertion disrupts the binding of a negatively acting transcription factor, thus causing a positive effect on transcription [22]. These investigators also showed that there was an ethnically distinct pattern of linkage disequilibrium between Hpa I apoc-i and apoe alleles, such that the H1 apoc-i allele was strongly associated with apoe o3, whereas the H2 apoc-i allele was strongly associated with apoe o2 and o4. This association was much stronger for European/Americans than for African /Americans. In the present study, we have attempted to confirm the presence of the same linkage disequilibrium of apoc- I and apoe alleles in a cohort of French/Canadians selected for a family study of familial combined hyperlipidemia (FCH). At the same time, we were interested in knowing whether the apoc-i H2 allele, which leads to increased apoc-i transcription in vitro, was associated with increased levels of circulating apoc- I. Our data in fact indicate that apoe gene polymorphisms have a much stronger influence than the apoc-i promoter polymorphism on plasma and lipoprotein concentrations of apoc-i. 2. Methods 2.1. Study subjects Data for the present study were obtained by analyzing stored (/70 8C) plasma and DNA samples, collected from 1998 to 2000, as part of an investigation of families with FCH. Twenty-nine families and 391 individuals were investigated in this original study. Probands with FCH were defined as individuals of French/Canadian descent with type IIB (combined) hyperlipidemia, having a plasma TG and LDL cholesterol (LDL-C) concentration greater than the 90th percentile for individuals of the same age and gender [23]. Two- or three-generation families had to have at least ten family members, with at least one individual other than the proband having hyperlipidemia (i.e. TG and/or LDL-C]/90th percentile for age and sex). Probands and confirming cases did not have a clinical or molecular diagnosis of familial hypercholesterolemia, severe obesity (BMI /35), diabetes (fasting glycemia ]/7.0 mmol/l), nor secondary hyperlipidemia due to thyroid, hepatic or renal disease. Nine families, representing samples from 178 individuals, were selected for measurement of apoc-i, apoe and apoc-iii. These families were chosen because they had a mixture of different apoe genotypes Lipid and apolipoprotein analyses Blood was obtained from subjects who had fasted for 12 h overnight. It was drawn under vacuum from an arm vein into tubes containing EDTA (final concentration: 1.5 mg/ml). Plasma was separated from blood cells by centrifugation (15 min, 3000 rpm, 4 8C). VLDL (db/ g/ml) were separated from 5 ml of fresh plasma by ultracentrifugation ( rpm, 10 h, 4 8C) in a 50.4 Ti rotor (Beckman Instruments, Palo Alto, CA), and bottom fractions (d /1.006 g/ml) were recovered quantitatively in a volume of 5 ml. Cholesterol and TG concentrations were determined enzymatically on an autoanalyzer (Cobas Mira, Roche). HDL cholesterol concentration was determined by assaying cholesterol in the supernatant, after precipitation of apob-containing lipoproteins from the d /1.006 g/ml fraction (1.0 ml) with heparin /manganese [24]. LDL-C levels were calculated as the difference between cholesterol in the d/ g/ml fraction and HDL cholesterol. ApoE phenotype was determined by isoelectric focusing gel electrophoresis followed by immunoblotting of whole plasma [25]. Plasma apoa-i and apob concentrations were determined by nephelometry (Behring Nephelometer 100 Analyzer). LDL apob concentration was the level of apob in the d/1.006 g/ml fraction. ApoC-I, apoc- III and apoe were measured with ELISAs developed in our laboratory [26 /28]. For the apoc-i assay, an immunopurified polyclonal goat anti-human apoc-i antibody (Biodesign, Kennebunk, ME) was employed for both capture and detection in a sandwich assay system. The apoc-i assay was calibrated with standard plasmas kindly provided by Dr Petar Alaupovic (Oklahoma Medical Research Foundation, Oklahoma City).

3 J.S. Cohn et al. / Atherosclerosis 169 (2003) 63/70 65 The intraassay and interassay CVs were 2.2 and 9.8%, respectively. ApoC-I, apoc-iii and apoe were measured in total plasma and in the VLDL (db/1.006 g/ml) fraction. ApoE, apoc-iii and apoc-i concentrations in the non-vldl (d/1.006 g/ml) fraction were calculated by difference DNA analyses DNA was extracted from white blood cells using an automated 340A DNA extractor (Applied Biosystems, Foster City, CA). Genotyping of the Hpa I apoc-i promoter region polymorphism was carried out according to Xu et al. [22]. In brief, 1 mg of genomic DNA was amplified by polymerase chain reaction using a GeneAmp PCR System 9600 (Perkin/Elmer Cetus, Norwalk, CT). The 450 bp final amplification products were digested with 5 U of the restriction enzyme HpaI (Gibco BRL, Rockville, MD) and separated by electrophoresis on 1.5% agarose gels for 30 min at 80 V. Resulting bands were stained with ethidium bromide and analyzed under UV light using an ALPHAEASE photoimager (Alpha Innotech, San Leandro, CA). ApoE genotyping was done by the restriction isotyping method of Hixson and Vernier [29]. In brief, 1 mg of genomic DNA was amplified by polymerase chain reaction. The 240 bp amplification products were digested with 5 U of the restriction enzyme HhaI (Gibco BRL) and separated on a 15% non denaturing polyacrilamide gel for 3 h at 70 V. The resulting bands were stained with ethidium bromide and analyzed under UV light Statistical analyses Data were analyzed using SIGMASTAT software (Jandel Corporation, San Rafael, CA). Statistically significant differences between genotypes were determined by one-way analysis of variance (ANOVA), or by Kruskal/Wallis one-way ANOVA on ranks, if data were not parametrically distributed. Linear regression analysis was used to calculate Pearson correlation coefficients (r), and hence assess the relationship between apoc-i and apoe concentrations. A P value of 0.05 or greater was taken to represent statistical significance. 3. Results Strong linkage disequilibrium was observed between apoc-i and apoe gene polymorphisms in individuals of French/Canadian descent (Table 1). There was a near perfect association of the apoc-i Hpa I-negative (H1) allele with the apoe3 isoform, and of the apoc-i Hpa I- positive (H2) allele with the apoe2 and E3 isoforms. Out of a total of 391subjects, 382 (97.7%) were found to have Table 1 ApoC-I HpaI genotype and apoe phenotype distribution in subjects recruited for family study ApoC-I Genotype H1/H1 H1/H2 H2/H2 Total ApoE Phenotype E2/E (100%) 12 E4/E (100%) 18 E3/E (96.9%) 2 (3.1%) 65 E3/E3 196 (98%) 4 (2%) E4/E3 1 (1.1%) 82 (96%) 2 (2.3%) 85 E4/E (100%) 11 Total Values represent number of subjects; values in parentheses are percentages of each apoe phenotype group. this association. Of the nine subjects with discordant results, four individuals were in three generations of a single family, and three others were in a second large 45 member family. In order to determine whether there was any association between the apoc-i promoter polymorphism and plasma apoc-i levels, frozen plasma and VLDL samples from nine families (i.e. 177 subjects) were analyzed for apoc-i, apoe and apoc-iii. Families were selected which had a mixture of different apoe genotypes. Half of these individuals were normolipidemic (n/89) and half were classified as being hyperlipidemic (n /88), based on their plasma TG being greater than 2.3 mmol/l and/or their LDL-C concentration being greater than 3.8 mmol/l. As shown in Table 2, the hyperlipidemic subjects tended to be older, of male gender, with higher BMIs. Twenty-eight subjects had an elevation in LDL-C concentration alone, 40 subjects had an elevation in TG concentration, and 20 subjects had an elevation in both. Table 2 Characteristics of study subjects Normolipidemic (n/89) Hyperlipidemic (n/88) Age (years) 419/2 499/2 Males/females 36/53 49/39 BMI (kg/m 2 ) 25.49/ /0.2 Plasma cholesterol 4.849/ /0.10 Plasma TG 1.359/ /0.12 LDL-C 2.829/ /0.09 HDL cholesterol 1.219/ /0.03 Values represent means9/s.e. for n individuals in each group. Lipid concentrations are in mmol/l and apo concentrations are in mg/ dl. Normolipidemic subjects had a plasma TG B/2.3 mmol/l and an LDL-C B/3.8 mmol/l. Hyperlipidemic subjects had a plasma TG /2.3 mmol/l and/or an LDL-C /3.8 mmol/l. By selection, the hyperlipidemic group had significantly higher mean TG, C and LDL-C levels (P B/0.001). They were also significantly older, had more males, had a significantly higher mean BMI and lower HDL cholesterol (P B/ 0.001).

4 66 J.S. Cohn et al. / Atherosclerosis 169 (2003) 63/70 Table 3 Plasma lipid and apo levels of normolipidemic subjects separated according to Hpa I apoc-i allele status H1/H1 (n/44) H1/H2 (n/37) H2/H2 (n/8) TG 1.309/ / /0.16 Cholesterol 4.869/ / /0.41 LDL-C 2.839/ / /0.38 LDL apob 849/3 839/3 809/11 HDL cholesterol 1.209/ / /0.10 ApoA-I 1529/4 1419/3 1439/8 Total apoc-i 10.59/ / /1.0 VLDL apoc-i 0.69/ / /0.2 d/1.006 apoc-i a 9.99/ / /1.0 Values represent means9/s.e. for n individuals in each group. Lipid concentrations are in mmol/l and apo concentrations are in mg/ dl. Normolipidemic subjects had a plasma TG B/2.3 mmol/l and LDL- C B/3.8 mmol/l. No statistically significant differences were found between the three groups by ANOVA. Normolipidemic subjects were divided according to apoc-i genotype and their mean plasma lipid and apo levels were compared (Table 3). Mean age and gender distribution of the three groups were not significantly different, and no significant differences were found between plasma lipid, apob or apoa-i concentrations. Plasma concentrations and lipoprotein distribution of apoc-i were also not significantly different between the three groups, providing no evidence for a link between apoc-i genotype and apoc-i levels (Table 3). Plasma concentrations and lipoprotein distribution of apoc-iii were also similar (data not shown). Total plasma apoe concentrations were, however, significantly higher in the H2/H2 group relative to the H1/H2 and H1/H1 groups: 6.19/0.6 vs. 4.69/0.2 and 4.29/0.2 mg/dl, P B/0.05, respectively. This reflected significantly higher non- VLDL (d/1.006 g/ml) apoe levels in the H2/H2 group relative to the H1/H2 and H1/H1 groups: 4.99/0.6 vs. 3.69/0.2 and 3.39/0.2 mg/dl, P B/0.05, respectively, and reflected the disproportionate number of apoe4/2 (n / 4) and apoe2/2 (n /2) compared with apoe4/4 (n/2) subjects in this small group. Normolipidemic subjects were subsequently separated according to apoe genotype (Table 4). The two subjects with an apoe2/2 genotype were included in the E3/2 group; the two subjects with an apoe4/4 genotype were included in the E4/3 group; the four subjects with an apoe4/2 group were omitted from this analysis. Mean ages for the three groups were not significantly different: 359/2, 429/4 and 429/2 years (apoe3/2 vs. apoe3/3 vs. apoe4/3), and gender distribution was also similar. As expected from previous studies of apoe genotype/ phenotype in relation to plasma lipid and apo levels [30,31], apoe3/2 subjects tended to have lower, while apoe4/3 individuals had higher LDL-C and LDL apob levels. ApoE4/3 subjects also tended to have higher plasma TG levels and lower HDL cholesterol and apoa- I levels (Table 4). As also predicted from previous studies [32/34], apoe3/2 subjects had higher and apoe4/ 3 subjects had lower levels of total and non-vldl apoe. Significantly, apoe3/2 subjects were found to have higher and apoe4/3 subjects to have lower levels of total and non-vldl apoc-i. In the normolipidemic group as a whole (n /89), total plasma apoc-i and apoe levels were significantly correlated (r /0.40, P B/ 0.001), as were non-vldl apoc-i and apoe levels (r/ 0.47, P B/0.001). These relationships were not entirely dependent on apoe genotype, however, since in the apoe3/3 group alone (n /45), total plasma apoc-i and apoe levels were significantly correlated (r /0.44, P B/ 0.01), as were non-vldl apoc-i and apoe levels (r/ 0.51, P B/0.001). Limited numbers of subjects in the other genotypic groups prevented meaningful correlational data to be obtained. Total and lipoprotein apoc- III levels were not significantly different in the apoe genotypic groups. In order to determine whether similar relationships could be observed in hyperlipidemic subjects, these individuals were divided according to their apoc-i genotype (Table 5). No significant differences were observed in total TG, total cholesterol, HDL cholesterol or apoa-i levels. LDL-C and apob levels were, however, significantly lower in H1/H2 versus H1/H1subjects, and lower still in H2/H2 versus H2/H1 subjects. Plasma and lipoprotein apoc-i levels were not significantly different between the three groups. VLDL apoc-i concentrations were, however, higher in H2/H2 individuals. As for normolipidemic subjects, total plasma apoe concentrations were significantly higher in the H2/H2 group relative to the H1/H2 and H1/H1 groups: 9.49/1.5 versus 6.69/0.5 and 5.69/0.2 mg/dl, respectively, P B/ 0.05, and apoe concentration in the non-vldl fraction was also significantly elevated in the H2/H2 group relative to the H1/H2 and H1/H1 groups: 5.99/0.9 versus 3.99/0.3 and 3.69/0.2 mg/dl, respectively, P B/ These differences were largely attributable to the presence in the H2/H2 group of four patients with an apoe2/2 genotype and type III hyperlipoproteinemia. They had appreciably higher plasma and lipoprotein apoe levels and also elevated levels of plasma and lipoprotein apoc-i, as shown in Table 6. This was consistent with the finding that subjects with an apoe3/2 genotype also tended to have increased apoe and apoc- I concentrations and reduced levels of LDL-C and apob relative to apoe3/3 and apoe4/3 individuals. Plasma and non-vldl apoc-i levels were, therefore, related to apoe genotype in hyperlipidemic subjects similar to that of normolipidemic subjects. VLDL apoc-i levels were also dependent on apoe genotype in hyperlipidemic subjects (Table 6), though this was not the case in normolipidemic subjects (Table 4). Plasma and lipoprotein levels of apoc-i were not entirely dependent on

5 J.S. Cohn et al. / Atherosclerosis 169 (2003) 63/70 67 Table 4 Plasma lipid and apo levels of normolipidemic subjects separated according to apoe genotype ApoE3/2 (n/22) ApoE3/3 (n/45) ApoE4/3 (n/18) Significance a TG 1.259/ / /0.09 P B/0.05 Cholesterol 4.469/ / /0.15 P B/0.05 LDL-C 2.409/ / /0.11 P B/0.001 LDL ApoB 699/4 859/2 979/3 P B/0.001 HDL cholesterol 1.239/ / /0.06 P B/0.01 ApoA-I 1509/4 1509/4 1329/5 P B/0.05 Total apoe 5.79/ / /0.2 P B/0.001 VLDL apoe 1.09/ / /0.1 ns d/1.006 apoe b 4.79/ / /0.2 P B/0.001 Total apoc-i 11.49/ / /0.5 P B/0.05 VLDL apoc-i 0.89/ / /0.1 ns d/1.006 apoc-i b 10.79/ / /0.6 P B/0.05 Total apoc-iii 13.39/ / /0.7 ns VLDL apoc-iii 4.09/ / /0.5 ns d/1.006 apoc-iii b 9.49/ / /0.6 ns Values represent means9/s.e. for n individuals in each group. Lipid concentrations are in mmol/l and apo concentrations are in mg/dl. Normolipidemic subjects had a plasma TG B/2.3 mmol/l and LDL-C B/3.8 mmol/l. a Statistically significant differences between genotypes were determined by one-way ANOVA, or by Kruskal/Wallis one-way ANOVA on ranks, if data were not parametrically distributed. ns, not significant. Table 5 Plasma lipid and apo levels of hyperlipidemic subjects separated according to Hpa I apoc-i allele status H1/H1 (n/34) H1/H2 (n/39) H2/H2 (n/15) Significance a TG 2.539/ / /0.27 ns Cholesterol 6.299/ / /0.28 ns LDL-C 3.889/ / /0.23 P B/0.05 LDL ApoB 1189/3 1089/3 919/5 P B/0.001 HDL 1.059/ / /0.06 ns cholesterol ApoA-I 1419/4 1549/6 1429/7 ns Total apoc-i 12.29/ / /1.0 ns VLDL apoc-i 1.89/ / /0.5 P B/0.05 d/1.006 apoc-i b 10.49/ / /1.1 ns Values represent means9/s.e. for n individuals in each group. Lipid concentrations are in mmol/l and apo concentrations are in mg/ dl. Hyperlipidemic subjects had a plasma TG /2.3 mmol/l and/or LDL-C /3.8 mmol/l. a Statistically significant differences between genotypes were determined by one-way ANOVA, or by Kruskal/Wallis one-way ANOVA on ranks, if data were not parametrically distributed. ns, not significant. b d/1.006 apoc-i represents that in the non-vldl fraction, which was determined by difference. apoe genotype, since in the apoe3/3 group (n /36), total plasma apoc-i and apoe levels were significantly correlated (r /0.47, P B/0.01), as were VLDL apoc-i and apoe levels (r/0.90, P B/0.001) and non-vldl apoc-i and apoe levels (r/0.44, P B/0.01). Limited numbers of subjects in the other genotypic groups prevented meaningful correlational data to be obtained. In the hyperlipidemic group as a whole (n/88), total plasma apoc-i and apoe levels were significantly correlated (r/0.37, P B/0.001), as were VLDL apoc-i and apoe levels (r /0.78, P B/0.001) and non-vldl apoc-i and apoe levels (r /0.27, P B/0.05). 4. Discussion The results of the present study confirm that there is strong linkage disequilibrium between apoc-i and apoe genes in French /Canadians, as observed previously in European/Americans [22]. In 382 out of 391 individuals studied (97.7%), the apoc-i Hpa I-negative (H1) allele was found in association with the apoe3 allele, while the apoc-i Hpa I-positive (H2) allele was found in association with the apoe2 and apoe4 alleles. As a result of this strong linkage between the apoc-i and apoe genes, French /Canadians were found to have plasma and lipoprotein apoc-i levels which were more strongly related to apoe genotype than to the Hpa I apoc-i polymorphism. In normolipidemic subjects, apoe3/2 genotype was associated with higher levels and apoe4/ 3 genotype was associated with lower levels of total and non-vldl apoc-i, while in hyperlipidemic subjects, apoe2/2 and apoe3/2 genotype was associated with increased total, VLDL and non-vldl apoc-i concentrations. Based on in vitro experiments with a reporter-gene assay, it has been proposed that by disrupting the binding of a negatively acting transcription factor, the Hpa I-positive (H2) allele is associated with a significant increase in apoc-i gene transcription [22]. The present results, however, show that any effect of this promoter

6 68 J.S. Cohn et al. / Atherosclerosis 169 (2003) 63/70 Table 6 Plasma lipid and apo levels in hyperlipidemic subjects separated according to apoe genotype ApoE2/2 (n/4) ApoE3/2 (n/17) ApoE4/2 (n/9) ApoE3/3 (n/36) ApoE4/3 (n/23) Significance a TG 3.689/ / / / /0.25 ns Cholesterol 7.529/ / / / /0.18 P B/0.05 LDL-C 3.309/ / / / /0.19 P B/0.05 LDL ApoB 789/13 979/5 949/6 1189/3 1129/4 P B/0.001 HDL cholesterol 1.149/ / / / /0.06 ns ApoA-I 1609/ /8 1379/6 1429/4 1539/7 ns Total apoe 17.19/ / / / /0.5 P B/0.001 VLDL apoe 6.39/ / / / /0.3 P B/0.001 d/1.006 apoe b 10.79/ / / / /0.3 P B/0.001 Total apoc-i 18.19/ / / / /0.4 P B/0.001 VLDL apoc-i 3.89/ / / / /0.3 P B/0.01 d/1.006 apoc-i b 14.39/ / / / /0.4 P B/0.05 Total apoc-iii 27.29/ / / / /1.7 ns VLDL apoc-iii 13.49/ / / / /1.4 ns d/1.006 apoc-iii b 13.89/ / / / /0.8 ns Values represent means9/s.e. for n individuals in each group. Lipid concentrations are in mmol/l and apo concentrations are in mg/dl. Hyperlipidemic subjects had a plasma TG /2.3 mmol/l and/or LDL-C /3.8 mmol/l. a Statistically significant differences between genotypes were determined by one-way ANOVA, or by Kruskal/Wallis one-way ANOVA on ranks, if data were not parametrically distributed; ns, not significant. b d/1.006 apo represents that in the non-vldl fraction, which was determined by difference. polymorphism on apoc-i gene transcription in vivo is not associated with a marked change in plasma or lipoprotein apoc-i levels. This is perhaps not surprising considering the number of factors, which can either affect plasma apoc-i production and/or catabolism. Some factors have been identified which can have a direct effect on apoc-i at the transcriptional level, e.g. LXR or RXR ligands [35], while others can indirectly influence plasma apoc-i levels (e.g. lipoprotein lipase, ABCA-I) by affecting VLDL and/or HDL metabolism [13,18]. Clearly, apoe is also an important determinant of plasma apoc-i levels, as reflected by the relationships which we have observed in both normolipidemic and hyperlipidemic subjects between apoe genotype and plasma/lipoprotein apoe and apoc-i levels (Tables 4 and 6). One explanation is that different apoe alleles are associated with different levels of apoc-i synthesis at a transcriptional level. Work by Taylor and colleagues has clearly demonstrated that the apoc-i and apoe genes reside in a gene cluster with apoc-ii and apoc-iv on chromosome 19 [36,37] and that these genes are regulated by common distal enhancer regions [38]. The close linkage disequilibrium between apoc-i and apoe genes thus makes it possible that common genetic polymorphisms affect the expression of both genes in a co-ordinated fashion. A second explanation, however, is that apoe isoforms in plasma can alter apoc-i concentrations through their differing effects on plasma lipoprotein metabolism. For example, apoe4 has increased binding affinity to TRL compared with apoe2, while apoe2 has a reduced ability to bind to lipoprotein receptors compared with apoe3 or apoe4 [39]. Asa result, individuals with an apoe2 isoform have higher apoe levels, and lower levels of LDL-C and apob than apoe4 carriers. Individuals with an apoe2 or apoe4 isoform also tend to have elevated levels of plasma TG [40]. These apoe-mediated effects on plasma lipoproteins could quite feasibly be the cause of secondary changes in plasma lipoprotein apoc-i concentration and distribution. Support for this concept is provided by our data showing that subjects with an apoe3/3 genotype (both normolipidemic and hyperlipidemic)) had apoc-i and apoe concentrations (i.e. plasma, VLDL and non- VLDL) which were significantly correlated, demonstrating that apoc-i levels were at least partly dependent on circulating apoe levels and not entirely dependent on apoe phenotype. It is interesting to note that normolipidemic subjects with an apoe3/2 genotype tended to have higher concentrations of plasma apoc-i, while subjects with an apoe4/3 genotype tended to have lower concentrations. These differences were mainly due to differences in non-vldl apoc-i rather than VLDL apoc-i levels (Table 4). They most probably reflected changes in HDL apoc-i rather than LDL or IDL apoc-i, since: (1) less than 10% of non-vldl apoc-i is associated with IDL or LDL in normolipidemic subjects [13,28], and (2) it is HDL apoe (non-apob-containing lipoprotein apoe), which is increased in patients with an apoe2/2 or E3/2 phenotype [33,34]. Based on the recent data of Gautier et al. [6], showing that apoc-i in HDL accounts for the ability of this lipoprotein to inhibit CETP activity, it is interesting to speculate that increased levels of HDL apoc-i in individuals having an apoe2 isoform

7 J.S. Cohn et al. / Atherosclerosis 169 (2003) 63/70 69 may reduce cholesteryl ester transfer activity, which could have a significant impact on reverse cholesterol transport in these subjects. In conclusion, the present study demonstrates that apoe gene polymorphisms have a much stronger influence than the apoc-i promoter polymorphism on plasma and lipoprotein concentrations of apoc-i. Since apoc-i appears to have several different lipid-regulating functions, further studies are warranted to define biological and genetic factors affecting the synthesis and secretion of this apo. Acknowledgements This study was funded by operating grants from the Heart and Stroke Foundation of Québec and the Canadian Institutes of Health (MT-14684). Financial support of our laboratory from Pfizer Canada in association with the Canadian Institutes of Health (DOP-40844) was also gratefully appreciated. References [1] Sehayek E, Eisenberg S. Mechanisms of inhibition by apolipoprotein C of apolipoprotein E-dependent cellular metabolism of human triglyceride-rich lipoproteins through the low density lipoprotein receptor pathway. J Biol Chem 1991;266:18259/67. [2] Clavey VS, Lestavel-Delattre S, Copin C, Bard JM, Fruchart JC. Modulation of lipoprotein B binding to the LDL receptor by exogenous lipids and apolipoproteins CI, CII, CIII, and E. Arterioscler Thromb Vasc Biol 1995;15:963/71. [3] Kowal RC, Herz J, Weisgraber KH, Mahley RW, Brown MS, Goldstein JL. Opposing effects of apolipoproteins E and C on lipoprotein binding to low density lipoprotein receptor-related protein. J Biol Chem 1990;265:10771/9. [4] Weisgraber KH, Mahley RW, Kowal RC, Herz J, Goldstein JL, Brown MS. Apolipoprotein C-I modulates the interaction of apolipoprotein E with beta-migrating very low density lipoproteins (beta-vldl) and inhibits binding of beta-vldl to low density lipoprotein receptor-related protein. J Biol Chem 1990;265:22453/9. [5] Jong MC, van Dijk KW, Dahlmans VE, van der Boom H, Kobayashi K, Oka K, Siest G, Chan L, Hofker MH, Havekes LM. Reversal of hyperlipidaemia in apolipoprotein C1 transgenic mice by adenovirus-mediated gene delivery of the low-densitylipoprotein receptor, but not by the very-low-density-lipoprotein receptor. Biochem J 1999;338:281/7. [6] Gautier T, Masson D, de Barros J-PP, Athias A, Gambert P, Aunis D, Metz-Boutigue M-H, Lagrost L. Human apolipoprotein C-I accounts for the ability of plasma high density lipoproteins to inhibit the cholesteryl ester transfer protein activity. J Biol Chem 2000;275:37504/9. [7] Buchko GW, Rozek A, Kanda P, Kennedy MA, Cushley RJ. Structural studies of a baboon (Papio sp.) plasma protein inhibitor of cholesteryl ester transferase. Protein Sci 2000;9:1548/58. [8] Jong MC, Voshol PJ, Muurling M, Dahlmans VE, Romijn JA, Pijl H, Havekes LM. Protection from obesity and insulin resistance in mice overexpressing human apolipoprotein C1. Diabetes 2001;50:2779/85. [9] Simonet WS, Bucay N, Pitas RE, Lauer SJ, Taylor JM. Multiple tissue-specific elements control the apolipoprotein E/C-I gene locus in transgenic mice. J Biol Chem 1991;266:8651 /4. [10] Shachter NS, Ebara T, Ramakrishnan R, Steiner G, Breslow JL, Ginsberg HN, Smith JD. Combined hyperlipidemia in transgenic mice overexpressing human apolipoprotein CI. J Clin Invest 1996;98:846/55. [11] Jong MC, Gijbels MJJ, Dahlmans VEH, van Gorp PJJ, Koopman S-J, Ponec M, Hofker MH, Havekes LM. Hyperlipidemia and cutaneous abnormalities in transgenic mice overexpressing human apolipoprotein C1. J Clin Invest 1998;101:145/52. [12] Holmquist L. Quantitation of human serum very low density apolipoproteins C-I, C-II, C-III and E by enzyme immunoassay. J Immunol Methods 1980;34:243/51. [13] Curry MD, McConathy WJ, Fesmire JD, Alaupovic P. Quantitative determination of apolipoproteins C-I and C-II in human plasma by separate electroimmunoassays. Clin Chem 1981;27:543/8. [14] Carlson LA, Holmquist L. Concentrations of apolipoproteins B, C-I, C-II, C-III and E in sera from normal men and their relation to serum lipoprotein levels. Clin Chim Acta 1982;124:163/78. [15] Riesen WF, Sturzenegger E. Enzyme-linked immunosorbent assay for apolipoprotein C-I. J Clin Chem Clin Biochem 1986;24:723/7. [16] Wahlberg G, Holmquist L, Walldius G, Annuzzi G. Effects of nicotinic acid on concentrations of serum apolipoproteins B, C-I, C-II, C-III and E in hyperlipidemic patients. Acta Med Scand 1988;224:319/27. [17] Bren ND, Rastogi A, Kottke B. Quantification of human plasma apolipoproteins C-I, C-II, and C-III by radioimmunoassays. Mayo Clin Proc 1993;68:657/64. [18] Alaupovic P, Schaefer EJ, McConathy WJ, Fesmire JD, Brewer HB, Jr. Plasma apolipoprotein concentrations in familial apolipoprotein A-I and A-II deficiency (Tangier disease). Metabolism 1981;30:805/9. [19] Annuzzi G, Holmquist L, Carlson LA. Concentrations of apolipoproteins B, C-I, C-II, C-III, E and lipids in serum and serum lipoproteins of normal subjects during alimentary lipaemia. Scand J Clin Lab Invest 1989;49:73/81. [20] Björkegren J, Karpe F, Milne RW, Hamsten A. Differences in apolipoprotein and lipid composition between human chylomicron remnants and very low density lipoproteins isolated from fasting and postprandial plasma. J Lipid Res 1998;39:1412/20. [21] Smit M, van de Kooil-Meijs E, Woudt LP, Havekes LM, Frants RR. Exact localization of the familial dysbetalipoproteinemia associated restriction site in the promoter region of the apoc1 gene. Biochem Biophys Res Commun 1988;152:1282/8. [22] Xu Y, Berglund L, Ramakrishnan R, Mayeux R, Ngai C, Holleran S, Tycko B, Leff T, Shachter NS. A common HPA I RFLP of apolipoprotein C-I increases gene transcription and exhibits an ethnically distinct pattern of linkage disequilibrium with the alleles of apolipoprotein E. J Lipid Res 1999;40:50/8. [23] Lipid Research Clinics Program Epidemiology Committee. Plasma lipid distribution in selected North American populations: the Lipid Research Clinics Program prevalence study. Circulation 1979;60:427/39. [24] Warnick GR, Albers JJ. A comprehensive evaluation of the heparin/manganese precipitation procedure for estimating high density lipoprotein cholesterol. J Lipid Res 1978;19:65/76. [25] Hill JS, Pritchard PH. Improved phenotyping of apolipoprotein E: application to population frequency distribution. Clin Chem 1990;36:1871/4. [26] Cohn JS, Tremblay M, Amiot M, Bouthillier D, Roy M, Genest J, Jr, Davignon J. Plasma concentration of apolipoprotein E in intermediate-sized remnant-like lipoproteins in normolipidemic and hyperlipidemic subjects. Arterioscler Thromb Vasc Biol 1996;16:149/59.

8 70 J.S. Cohn et al. / Atherosclerosis 169 (2003) 63/70 [27] Fredenrich A, Giroux L-M, Tremblay M, Krimbou L, Davignon J, Cohn JS. Plasma lipoprotein distribution of apoc-iii in normolipidemic and hyperlipidemic subjects: comparison of the apoc-iii to apoe ratio in different lipoprotein fractions. J Lipid Res 1997;38:1421/32. [28] Cohn JS, Tremblay M, Batal R, Jacques H, Veilleux L, Rodriguez C, Bernier L, Mamer O, Davignon J. Plasma kinetics of VLDL and HDL apoc-i in normolipidemic and hypertriglyceridemic subjects. J Lipid Res 2002;43:1680/7. [29] Hixson JE, Vernier DT. Restriction isotyping of human apolipoproteine by gene amplification and cleavage with HhaI. J Lipid Res 1990;31:545/8. [30] Davignon J, Gregg RE, Sing CF. Apolipoprotein E polymorphism and atherosclerosis. Arteriosclerosis 1988;8:1/21. [31] Schaefer EJ, Lamon-Fava S, Johnson S, Ordovas JM, Schaefer MM, Castelli WP, Wilson PWF. Effects of gender and menopausal status on the association of apolipoprotein E phenotype with plasma lipoprotein levels. Results from the Framingham Offspring Study. Arterioscler Thromb 1994;14:1105/13. [32] Smit M, de Knif P, Rosseneu M, Bury J, Klasen E, Frants R, Havekes L. Apolipoprotein E polymorphism in the Netherlands and its effect on plasma lipid and apolipoprotein levels. Hum Genet 1988;80:287/92. [33] Luc G, Ducimetiere P, Bard J-M, Arvelier D, Evans A, Cambien F, Fruchart J-C, Fievet C. Distribution of apolipoprotein E between apo B- and non apo B-containing lipoproteins according to apo E phenotype. Atherosclerosis 1997;131:257/62. [34] Schiele F, Vincent-Viry M, Starck M, Beaud B, Hennache G, Siest G, Visvikis S, Herbeth B. Apolipoprotein E in apolipoprotein B (apo B)- and non-apo B-containing lipoproteins in 3523 participants in the Stanislas Cohort: biological variation and genotypespecific reference limits. Clin Chem 2002;48:291/300. [35] Mak PA, Laffitte BA, Desrumaux C, Joseph SB, Curtiss LK, Mangelsdorf DJ, Tontonoz P, Edwards PA. Regulated expression of the apolipoprotein E/C-I/C-IV/C-II gene cluster in murine and human macrophages; a critical role for the nuclear receptors LXRalpha and LXRbeta, J Biol Chem 2002;277:31900/8. [36] Allan CM, Walker D, Taylor JM. Evolutionary duplication of a hepatic control region in the human apolipoprotein E gene locus. Identification of a second region that confers high level and liverspecific expression of the human apolipoprotein E gene in transgenic mice. J Biol Chem 1995;270:26278/81. [37] Allan CM, Walker D, Segrest JP, Taylor JM. Identification and characterization of a new human gene (APOC4) in the apolipoprotein E, C-I, and C-II gene locus. Genomics 1995;28:291/300. [38] Shih SJ, Allan C, Grehan S, Tse E, Moran C, Taylor JM. Duplicated downstream enhancers control expression of the human apolipoprotein E gene in macrophages and adipose tissue. J Biol Chem 2000;275:31567/72. [39] Mahley RW, Rall SC, Jr. Type III hyperlipoproteinemia (dysbetalipoproteinemia): the role of apolipoprotein E in normal and abnormal lipoprotein metabolism. In: Scriver CR, Beaudet AL, Sly WS, Valle D, editors. The metabolic basis of inherited disease. New York, NY: McGraw-Hill Publishing Co, 1989:1195/213. [40] Dallongeville J, Lussier-Cacan S, Davignon J. Modulation of plasma triglyceride levels by apoe phenotype: a meta-analysis. J Lipid Res 1992;33:447/54.