Introduction

Increased concentrations of plasma lipids (total cholesterol (TC) and triglycerides (TG) are a risk factor for myocardial infarction development,1 and are known to be modified by both genetic and environmental factors.

Apolipoprotein E (APOE) is a structural component of TG-rich lipoproteins and it serves as a ligand for the lipoprotein receptors. Three common APOE isoforms (apoE4 (Cys112>Arg), apoE3 and apoE2 (Arg158>Cys)) are known. The APOE4 allele is associated with high and the APOE2 allele with low levels of TC.2

Apolipoprotein A5 (APOA5) is located on TG-rich and on HDL-particles and plays a role in lipoprotein lipase activation.3 In the APOA5 gene, variants T-1131>C and Ser19>Trp are associated with differences in TG levels.4, 5

The aim of our study was to evaluate if there is any interaction between the APOE and APOA5 gene variations and plasma lipids in a large Caucasian population.

Materials and methods

Subjects

The 1168 men and 1332 women (response rate of 84%) represent a 3-year cohort of a selected 1% Czech Caucasian population sample. The individuals were recruited from nine districts in 1997–1998 and re-invited in 2000–2001 according to the WHO protocol (MONICA Project. Manual WHO/MNC 82.2, November 1983). Written informed consent was obtained from the study participants and the local ethics committee approved the design of the study.

Genetic and biochemical analysis

DNA was isolated6 and APOE7 and APOA58 were genotyped as described before. The lipoprotein parameters were measured by the WHO Regional Lipid Reference Centre, IKEM, Prague on a Roche COBAS–MIRA autoanalyser, using reagents from Boehringer Mannheim Diagnostics and Hoffmann-La Roche.

Statistical analysis

Analysis of variance for repeated measures was used for the statistical analysis. The individuals with APOE4/E2 genotypes and individuals, where not all genotypes were available' were excluded. According to APOA5 genotypes, the individuals were separated into two subgroups; the carriers of the TT-1131/SerSer19 haplotype and the carriers of at least one allele C-1131 or Trp19. Within each group, the individuals were further separated as APOE2 carriers, APOE3E3 homozygotes and APOE4 carriers. Values are given like mean (SD).

Results

Study population

The basic characteristics are summarized in Table 1. The distributions of APOE alleles (male/female: APOE2=7.6/7.7%, APOE3=82.6/82.5%, APOE4=9.8/9.8%) and APOA58 genotypes are similar to the frequencies described for other Caucasians.9 The distribution of combinations of APOE and APOA5 genotypes is summarized in Table 2.

Table 1 Basic characteristics of the analyzed individuals
Full size table
Table 2 Distribution of the APOA5 haplotypes and the groups of APOE genotypes in the Czech population
Full size table

Effects of the individual APOE and APOA5 genotypes

Both in male subjects and female subjects, APOE2 carriers have the lower levels of TC and APOE4 carriers have higher levels of TC in comparison with the APOE3E3 homozygotes (Table 3a; P<0.01).

Table 3 Plasma levels of total cholesterol (a) and triglycerides (b) according the APOA5 haplotypes and the groups of APOE genotypes in the Czech population
Full size table

Also the APOA5 gene has the usual effect on TG levels – individuals carrying the Trp19 and C-1131 alleles have higher levels of TG (P<0.001) than the others.8

Interaction between APOE and APOA5 genotypes and plasma lipid levels

Total cholesterol and TG levels have been affected significantly by combinations of APOE genotypes and APOA5 haplotypes in female subjects (P<0.05 for interaction, F=3.95, d.f. 2; 1324). The effect of APOE was detectable only in individuals with the most common haplotype T-1131T/Ser19Ser (P<0.001 for trend; Table 3a). In individuals with at least one APOA5 allele C-1131 and/or Trp19, APOE has no effect on TC levels (Table 3b). Additionally, if APOE4 allele is present; there was no effect of APOA5 on TG levels (Table 3b). TG levels are elevated in carriers of at least one APOA5 allele C-1131 and/or Trp19 only if APOE2 or APOE3 alleles are present (P<0.01; P<0.05 for interaction, F=4.02, d.f. 2; 1338).

Combinations of APOA5 and APOE variants showed no significant effects on the plasma lipids in male subjects.

Plasma TC were significantly lower in APOE2 allele carriers in comparison with the APOE4 carriers (P<0.05) regardless of the APOA5 haplotype (Table 3a). Similarly, the carriers of the less common APOA5 alleles have elevated TG levels in comparison to others.8 We have detected no significant differences in TG levels after dividing these subgroups according the APOE genotypes despite the fact that a non-significant trend, similar to the results observed in female, is present (Table 3b).

Discussion

The final levels of plasma lipids are under polygenic control, with more variants in a couple of genes contributing in the total effect. These genes will affect plasma lipids together with external/environmental (smoking, dietary habits, menopausal status, physical activity, etc.) factors. This makes the discovery of gene–gene–environment interaction very complicated.

In our study, we have detected that the effect of interactions between variants in the APOA5 and APOE genes on plasma lipids is more pronounced in female than in male subjects.

In women, carriers of at least one APOA5 allele C-1131 or Trp19, APOE polymorphism has no effect on plasma TC. APOE exhibit the effect on TC only in individuals with commonest APOA5 haplotype T-1131T/Ser19Ser. Further, the effect of APOA5 variants on TG levels was detectable in individuals with APOE2E2, APOE3E3 or APOE3E2 genotypes, but not if the APOE4 was present. In men, a similar, non-significant trend was visible.

The number of published studies analyzing gene–gene interaction in connection to atherosclerosis is not high, but some interesting examples could be selected.

For example, an interaction between APOE4 allele and peroxisome proliferator-activated receptor (PPAR; nuclear transcription factor) CT genotype and plasma TC and the risk of coronary heart disease was detected.10 Furthermore, the combinations of the APOE and cholesterol ester transfer protein alleles have significant effect on plasma HDL-cholesterol levels.11

Finally, the APOE–APOA5 interaction was analyzed in hypertriglyceridemic individuals. Schaefer et al12 found six APOE2/E2-APOA5Trp19 carriers out of 170 individuals with TG levels over 2.3 mmol/l and this combination was not detected in healthy normolipidemic individuals. We did not confirm this interaction, but in 111 patients with extreme TG levels (>10 mmol/l) the combination APOE4–APOA5Trp19 was overrepresented (P<0.005) in comparison with the healthy population (N=2559).13

Nothing is known about the possible mechanism of interaction between APOE and APOA5 variants in genetic determination of plasma lipids. APOE negotiates the interaction of lipoprotein particles with lipoprotein particles and APOA5 plays a role in the activation/stabilisation of lipoprotein lipase, but their role in lipid metabolism is still not completely understood. As both proteins are located on TG-rich particles, we can speculate about the possibility that lipoprotein particles in individuals with common APOA5 haplotype have a slightly delayed half-life in plasma and thus the effect of APOE variants could be more efficient. Why this effect is more manifested in women than in men remains questionable.

It is evident that the effect of APOE genotype on lipid parameters may be modified by other genes. We conclude that there could be a sex-specific interaction between variants in APOE and APOA5 genes, which may play a role in genetic determination of plasma levels of lipids in female subjects, but not in male subjects. The mechanism of the interaction needs to be analyzed in detail.