Atherogenic and anti-atherogenic lipoprotein functions

Although the numerous atheroprotective effects ofn-3 PUFAs are well known, the effects of these fatty acids on the underlying mechanisms of atherosclerosis have been less studied in humans (196). Lipid retention in the arterial intima, in which the entrapment of lipoproteins by arterial proteoglycans plays a central role, is considered the initiating event in atherosclerosis (7). Previously, Jones et al. (202)

85 observed slightly decreased binding of isolated LDL to one of the arterial proteoglycans, biglycan, after high-oleic canola oil and corn/safflower oil intakes whereas DHA-enriched high-oleic canola oil failed to induce any change. In the present study, instead of isolated LDL particles we used whole plasma. We were, therefore, able to assess the effect of all plasma lipoproteins on cholesterol accumulation to proteoglycans. Intake of CSO was found to decrease the binding of lipoproteins to aortic proteoglycans, whereas fish intake had no effect on lipoprotein-proteoglycan interactions.

To study further which lipoprotein particles were responsible for these results, we analyzed the correlations between lipoproteins and their interactions with proteoglycans. We found strong correlations between changes in the concentrations of apoB, total cholesterol, LDL cholesterol and apoB-containing lipoprotein subclasses and changes in the binding of lipoproteins to proteoglycans. Changes in concentrations of IDL and LDL subclasses were found to have the strongest correlation with lipoprotein-proteoglycan interactions. In contrast, HDL particles and lipoprotein-proteoglycan interactions were not correlated, except for HDL³ (data not shown). Similarly, in the study by Olin-Lewis et al. (249), apoB-containing lipoproteins were found to bind to biglycan with various degrees of affinity. LDL tended to have the greatest binding affinity followed by IDL and VLDL. Binding affinity of LDL subclasses also increased with the decreasing diameter of the particles. These findings highlight the importance of apoB-containing lipoproteins in the accumulation of cholesterol within the artery wall.

A possible mechanism for the changes in lipoprotein-proteoglycan interactions could be compositional changes in apoB-containing particles, especially LDL, which affect the conformation of apoB-100 (7). Conformational changes in apoB-100 could expose more proteoglycan binding sites on the particle surface. However, as LDL cholesterol concentration and IDL particle concentrations were found to decrease in the CSO group in the present study, there were fewer apoB-containing particles in plasma to interact with proteoglycans. Thus, our results together with previous findings suggest that the effects of dietary fatty acids on lipoprotein-proteoglycan interactions may be mediated by fatty acids other than EPA and DHA. Future research is needed to further explore these relations.

The prolonged exposure of lipoproteins entrapped by proteoglycans to local enzymes and other factors within the artery wall makes these particles susceptible to several modifications (7,8). These modifications, especially in the LDL particle surface, lead to the loss of particle stability and makes LDL particles prone to aggregation. Very little is known about the effects of diet on LDL aggregation in humans, but a healthy Nordic diet has been found to decrease LDL aggregation (182).

These effects were found to be associated with MUFA and PUFA intakes. In the present study, intake ofn-3 PUFAs from either fish or CSO had no effect on LDL aggregation. However, individual differences in LDL aggregation were large, as reported also by Ruuth et al. (182). Changes in LDL aggregation have been found to be related to lipids on the LDL surface. A decrease in LDL sphingomyelins has been

86

associated with reduced LDL aggregation (182). The sphingomyelin content of plasma has been shown to be genetically regulated, which could, at least in part, explain the inter-individual variance found in LDL aggregation (250).

Modified LDL particles can trigger several cellular responses within the arterial intima, which further increase the accumulation of lipids in the artery wall and accelerate the progression of atherosclerosis (7,175). One of these responses is secretion of IL-8 from endothelial cells, which can induce the chemotactic recruitment of monocytes into the arterial intima. The release of IL-8 from endothelial cells can be stimulated by modified LDL particles (175,251). In these previous studies, LDL lipid composition was modulated enzymatically and increased content of free fatty acids and lysophosphatidylcholine in LDL was found to enhance IL-8 release. There are no previous intervention studies investigating the effects of diet on the ability of LDL to release of IL-8. However, Padro et al. (252) found no changes among lipid species of LDL after intake of milk enriched with long-chainn-3 PUFAs (0.375 g/250 ml milk). However, they observed changes in the PC36:5/lysoPC16:0 ratio and in long-chainn-3 PUFA content in CE and PC, which may reduce the inflammatory activity of LDL particles. We found no effect of either fish or CSO intakes on the pro-inflammatory properties of LDL.

HDL has the ability to counteract the pathogenic events of atherosclerosis via the RCT pathway, which promotes the removal of cholesterol from macrophages in atherosclerotic plaques (185). Cholesterol efflux is a critical part of the RCT pathway.

Despite the importance of cholesterol efflux in attenuating the progression of atherosclerosis, the effects of diet on this process have not been thoroughly investigated. Previous human intervention studies have found different dietary fatty acids to improve cholesterol efflux capacity (203–208). Montoya et al. (203) found bothn-3 andn-6 PUFA diets to increase cholesterol efflux capacity from Fu5AH cells to serum. Furthermore, Tanaka et al. (207) found EPA supplementation with a high dose (1.8 g/day) to enhance cholesterol efflux capacity of HDL from THP-1 macrophages in dyslipidemic subjects.

In contrast to these findings, we observed no change in cholesterol efflux capacity from human primary macrophages to HDL in the fatty fish group. However, in experimental models, direct exposure of human primary or THP-1 macrophages to EPA has been found to reduce the cholesterol efflux capacity (253,254), but exposure of macrophages to both EPA and DHA has been observed to increase apoA-1-mediated cholesterol efflux (255). The discrepancies in these results may be related to the different study settings and variation in the donor and acceptor cells studied.

In the present study, the intake of EPA in the fatty fish group was significantly lower (approximately 0.5 g/day) than in the study by Tanaka et al. (207). Furthermore, the sample size (n=16) in our cholesterol efflux assay may have been too small to detect any changes.

87 Concentrations of apoE and SAA and their correlations with HDL particles

SAA and apoE have been found to be significant contributors to atherosclerosis (194,256). Concentrations of SAA are increased in a variety of disease states, and it has been shown to have pro-atherogenic effects (194). ApoE is best recognized for its anti-atherogenic functions, such as contributing to cholesterol efflux (256). We found no changes in the plasma apoE concentration. Interestingly, the serum concentration of SAA decreased in the CSO and lean fish groups, although we did not observe changes in other inflammatory markers among the groups, as previously reported (214). In line with our results, a diet enriched with linseed oil was previously found to decrease the concentration of SAA in dyslipidemic subjects (257).

We observed a decrease in SAA in the lean fish group but not in the fatty fish group. This may be related to the effects of DHA, which has been found to increase the expression of SAA in a cell culture study (258). Furthermore, amino acids and bioactive peptides in the fish may exert anti-inflammatory effects (259). One example of these potential compounds with anti-inflammatory activity is taurine, which has been found to decrease hs-CRP (high-sensitivity C-reactive protein) concentration after supplementation (3 g/day) for 8 weeks (260). Taurine is found in higher content in lean fish than in fatty fish (261,262). Further studies are needed to investigate the anti-inflammatory effects of lean fish and the potential bioactive compounds it contains. Of note, after we excluded the individuals with hs-CRP > 10 mg/l at baseline or at the end of the study (n=3), we found no differences between the groups in the pairwise comparison (data not shown).

Both SAA and apoE can also mediate HDL-proteoglycan interactions (194,195). In order to clarify this relationship between HDL and proteoglycan interactions in our study, we first investigated the correlations between SAA and apoE and HDL particles. We found that SAA but not apoE tended to correlate with HDL particles, especially at baseline. Correlations at the end of the study were found only between SAA and HDL³, medium and small HDL particles. We also examined the correlations between HDL particles and proteoglycan binding, but found no correlations, except for HDL³ (data now shown). However, HDL³ was also found to correlate with LDL subclasses (data not shown). Thus, our results suggest that HDL particles do not significantly mediate lipoprotein-proteoglycan interactions, and the relation of HDL³ to these interactions may be a more general indication of a dyslipidemic profile that also includes small HDL particles (192).

Although we did not observe significant interactions between HDL and arterial proteoglycans, this interplay may interfere with the formation of LDL-proteoglycan complexes. Umaerus et al. (195) found HDL2 to inhibit the formation of these complexes more efficiently than HDL³, which was suggested to be related to the higher apoE content in large HDL particles than in small HDL particles. However, the retention of HDL in the arterial intima can also be deleterious. Entrapped HDL cannot promote cholesterol efflux and is also exposed to same modifications as other retained lipoproteins (185). Further studies, especially using isolated HDL particles,

88

are needed to elucidate the potential of HDL to interact with arterial proteoglycans and the role of these interactions in the progression of atherosclerosis.