Subclasses

Methods used for determing lipoprotein subclasses

Lipoproteins can be fractioned according to physiochemical properties and chemical composition with several techniques, such as ultracentrifugation (UC), gradient gel electrophoresis (GGE) and nuclear magnetic resonance (NMR) spectroscopy (124,125). Sequential UC separates lipoproteins according to their density. HDL particles can be fractioned into two classes: large and buoyant HDL² and small and dense HDL³. Similarly, LDL particles can be categorized by UC into large and small particles, more specifically into three or four major subclasses: large and buoyant (LDL I), intermediate (LDL II) and small and dense (LDL III). Sometimes very small LDL (LDL IV) is also separated.

Different subclasses of LDL and HDL can also be fractioned with GGE in which the separation is based on the size and charge of lipoproteins (124,125). LDL can be divided into four major subclasses (I–IV) and further into two subtypes for LDL III and IV. Different LDL subtypes are sometimes also referred as type A in which large, buoyant are predominant and type B in which small and dense are predominant.

Two subfractions of VLDL and IDL have also been detected by ultracentrifugation and gel electrophoresis: large, more buoyant VLDL1 and small, dense VLDL2 and large, triglyceride-rich IDL-I and smaller, cholesterol-rich IDL-II (126–128). HDL is separated by GGE into HDL^2b, HDL^2a, HDL^3a, HDL^3b and HDL^3c which decrease in size, respectively. With 2-dimensional gel electrophoresis HDL particles can be separated by size in vertical dimension and charge in the horizontal dimension into more than 10 distinct subclasses (124).

UC and GGE require physical separation of the particles, whereas NMR spectroscopy is based on the magnetic properties of particles, specifically NMR signals of the terminal lipid methyl group protons (124,129). Furthermore, NMR spectroscopy enables the quantification of the complete lipoprotein profile, including

38

chylomicrons, several VLDL subclasses and IDL particles (129). With NMR spectroscopy up to 26 subclasses of HDL and 47 subclasses of LDL, VLDL and chylomicrons have been detected, but usually these subclasses are grouped into a few main categories (124). Each of these methods is based on different physiochemical properties, and there is overlap between these subclasses separated with different methods, which make the comparison of results difficult. Furthermore, there is no generally agreed reference method to measure lipoprotein subclasses.

Factors influencing the lipoprotein subclass profile

Several factors affect the lipoprotein subclass profile, such as gender (130,131), age (130), ethnicity (131), physical activity (132), use of statins (133) and alcohol consumption (134,135). Lipoprotein subclasses are also regulated by genetic factors (136,137), but diet may also have important effects. A dietary pattern that includes sweets, hamburgers, pizzas and salty snacks have been associated with smaller particle size of LDL, increased particle size of VLDL and increased concentration of small HDL particles (138). In contrast, Mediterranean diet enriched with nuts has been found to decrease the concentrations of large VLDL, medium-small LDL and very small LDL as well as increase the concentrations of large LDL, total HDL particles and mean particle size of LDL (139).

Although the beneficial effects ofn-3 PUFA on concentrations of plasma TG and HDL cholesterol is well-established (19), the effects of these fatty acids on lipoprotein subclasses have been less studied. In dietary intervention studies, fish intake has been found to decrease the size and concentration of VLDL particles and increase the size and concentration of HDL particles as compared with control diets (140–147) (Table 6). Furthermore, intake of totaln-3 PUFAs and a dietary pattern high in fish have been associated with lower concentrations of large and medium VLDL particles and smaller particle size of VLDL (148,149). Furthermore, intake of totaln-3 PUFAs have been positively associated with the average size of HDL particles. The effects of ALA intake on lipoprotein subclass profile has been less studied and the current evidence is inconsistent (45,52,150–152) (Table 7).

39

Table 6. Overview of intervention studies investigating the effects of fish intake on lipoprotein subclasses in order of publication year Continued on next page.ReferenceSubjectsStudy designStudy dietsAnalytical methodMain results Lacaille et al. 2000 (145)

n=11, M Normolipidemic individuals Age: 19–27 years Randomized, controlled trial, cross-over 1) Lean fish (cod and sole, 70-75% of daily protein) 2) Non-fish (beef, pork, veal, eggs, milk and milk products) for 4 weeks (5-week washout) Dextran-sulfate precipitation: HDL2and HDL3

HDL2 Ratio of HDL2 and HDL3 after lean fish intake Beauchesne- Rondeau et al. 2003 (144)

n=18, M Hypercholesterolemic individuals Age: 21–73 years Randomized, controlled trial, cross-over 1) Lean fish (pollack, cod, sole and haddock, 69% of daily protein) 2) Lean beef 3) Poultry for 26 days (6-week washout) Dextran-sulfate precipitation: HDL2and HDL3

HDL2 Ratio of HDL2 and HDL3 after lean fish intake Li et al. 2004 (140)n=22, M+ F Age: > 40 years Mildly hypercholesterolemic individuals

Parallel interventionNational Cholesterol Education Program (NCEP) diet + 1) High-fish (tuna, salmon and sole 8 times/week) 2) Low-fish (fish 2 times/week + lean poultry) for 24 weeks Dextran-sulfate precipitation: HDL2and HDL3 NMR

HDL2 and HDL3 after both fish diets Concentration of medium and small VLDL after high-fish diet Concentration of IDL and large HDL particles and HDL particle size after low-fish diet Ouellet et al. 2008 (153)n=19, M + F Overweight or obese individuals with impaired glucose metabolism Age: 40–65 years

Randomized controlled trial, cross-over 1) Cod (58–68% of daily protein) 2) Non-fish (lean beef, pork, veal, eggs, milk and milk products) + cod liver oil for 4 weeks (4-week washout) Dextran-sulfate precipitation: HDL2and HDL3

HDL2 and HDL3 Lindqvist et al. 2009 (141)

n=35, M Overweight individuals Age: 35–60 years Randomized controlled trial, cross-over 1) Herring (5 meals/week) 2) Control (lean pork and chicken) for 6 weeks (12-week washout) Dextran-sulfate precipitation: HDL2and HDL3

HDL2 after herring intake

40

Table 6. (continued) F, female; M, male; NMR, nuclear magnetic resonance spectroscopy.

ReferenceSubjectsStudy designStudy dietsAnalytical methodMain results Lankinen et al. 2014 (142) n=105, M + F Individuals with impaired glucose metabolism and features of metabolic syndrome Age: 40–70 years Randomized controlled trial, parallel 1) Fatty fish (3 meals/ week +wholegrain and bilberries) 2) Habitual fish consumption (+ wholegrain) 3) Control (1 fish meal/week) for 12 weeks

NMRLarge HDL particles HDL particle size after increased fish intake Erkkilä et al. 2014 (143)

n=33, M + F Individuals with coronary heart disease Age: < 70 years Randomized, controlled trial, parallel 1) Fatty fish (4 meals/week) 2) Lean fish (4 meals/week) 3) Control (1 fish meal/week) for 8 weeks

NMRHDL particle size after fatty fish intake Aadland et al. 2015 (146)

n=20, M + F Healthy individuals Age: 18–65 years Randomized, controlled trial, cross-over 1) Seafood diet (2 meals/day, cod, pollock, saithe and scallops) 2) Non-seafood diet (lean meat and dairy + cod liver oil) for 4 weeks (5-week washout)

NMRVLDL particle size after lean fish intake Raatz et al. 2016 (147)

n=19, M + F Overweight individuals Age: 40–65 years Randomized trial, cross-over1) 90 g of salmon 2 meals/week 2) 180 g of salmon 2 meals/week 3) 270 g of salmon 2 meals/week for 4 weeks (4- to 8-week washout)

NMRMedium HDL, small HDL after the lowest dose Medium HDL, small HDL after the highest dose VLDL particle size, LDL particle size after every dose Rundblad et al. 2018 (84)

n=36, M + F Healthy individuals Age: 18–70 years Randomized, controlled trial, parallel 1) Fish (3 meals/week, cod, salmon, mackerel) 2) Krill oil (4 g of krill oil/day, 1 fish meal/week) 3) Control (1 fish meal/week) for 8 weeks NMRVery small VLDL after krill oil intake HDL3 after krill oil and fish intakes

41

Table 7. Overview of intervention studies investigating the effects of ALA intake on lipoprotein subclasses in order of publication year ALA, -linolenic acid; F, female; LA, linoleic acid; M, male; NMR, nuclear magnetic resonance spectroscopy; PUFA, polyunsaturated fatty acid.

ReferenceSubjectsStudy designStudy dietsAnalytical methodMain results Kratz et al. 2002 (150)n=56, M + F Healthy individuals Age: 18–43 years Randomized intervention, parallel 1) Canola oil (2.5 E% ofn-3 PUFAs) 2) Sunflower oil 3) Olive oil for 4 weeks

Gradient gel electrophoresisLDL particle size after all diets Goyens et al. 2005 (151)n=54, M + F Healthy individuals Age: 18–65 years

Randomized, double-blind, controlled trial, parallel Fish consumption not allowed during intervention 1) High-ALA diet (1.1 E% of ALA) 2) Low-LA diet (0.4 E% of ALA) 3) Control (0.4 E% of ALA) for 6 weeks

NMRTotal and small VLDL after high-ALA diet Wilkinson et al. 2005 (45)n=57, M Individuals with atherogenic lipoprotein phenotype Age: 35–60 years

Randomized, single-blind intervention, parallel 1) Flaxseed oil and spread (15 g of ALA/day) 2) Sunflower and fish oil 3) Sunflower oil for 12 weeks

Density gradient ultracentrifugationNo significant effects on subclasses after flaxseed oil consumption Small LDL, HDL2 after sunflower/fish oil diet Harper et al. 2006 (152)n=56, M + F Overweight individuals Average age: 51 years

Randomized, double-blind, controlled trial, parallel 1) Flaxseed oil capsules (3 g of ALA/day) 2) Olive oil capsules for 26 weeks

Density gradient ultracentrifugationNo significant effects on subclasses Dodin et al. 2008 (52)n=179, F Healthy, menopausal Age: 49–65 years

Randomized, double-blind, controlled trial, parallel 1) 40 g of flaxseeds (9 g of ALA/day) 2) 40 g of wheat germ for 12 months Gradient gel electrophoresisNo significant effect on LDL particle size

42

Lipoprotein subclasses and the risk of CVD

Different lipoprotein subclasses have been found to be related to varying degrees with the risk of CVD (154,155). Large HDL particles may be more protective against CVD than small HDL particles (154,156). Differences in the protective abilities of different HDL subclasses may be related to the different biological activities of each subclass (120). Furthermore, certain pathologic conditions, such as acute infections, CVD or diabetes, alter the structure and composition of HDL, which leads to the loss of its beneficial activities (122). Such altered HDL particles are known as dysfunctional HDL. Like HDL particles, also LDL subclasses vary in their impact on CVD risk (157). Small LDL particles have been found to be more atherogenic than large LDL particles (158,159). Small LDL particles predict the risk of future coronary heart disease (CHD) events also in individuals with low risk of CHD based on their LDL cholesterol concentration (158). Small LDL particles contain less cholesterol and, therefore, an increase in the concentration of these particles may not be reflected in LDL cholesterol concentration. These LDL particles are considered to be more atherogenic than large particles due their slower clearance from circulation, increased affinity to arterial proteoglycans, vulnerability to oxidation and potentially increased transport into the subendothelial space of arteries (160).

Although VLDL and IDL have normally smaller concentrations in plasma, they have been found to possess atherogenic properties (161,162). Increased concentrations of large VLDL particles have been observed in individuals with metabolic syndrome (163) and type 2 diabetes (164). Large VLDL particles have also been found to be precursors for small, dense LDL particles (165). Less is known about atherogenicity of IDL subspecies. However, smaller, cholesterol-rich (IDL-II), but not large, triglyceride-rich IDL particles (IDL-I) have been found to bind to aortic proteoglycans inin vitro models (127).

Abnormal lipids and lipoproteins commonly appear together during metabolic perturbations such as obesity, type 2 diabetes and metabolic syndrome (166). This cluster of abnormalities is usually characterized by elevated plasma TG concentration, low concentration of HDL cholesterol and the presence of small, dense LDL particles (Figure 5). Elevated concentrations of other apoB-containing lipoproteins and changes in the lipid species of lipoproteins are also observed. The terms used for this lipid profile are “atherogenic lipoprotein phenotype”, “lipid triad” and “atherogenic dyslipidemia”. Dysregulation of lipoprotein metabolism has a key role in the development of atherosclerosis (166).

43 Figure 5. Features of the atherogenic lipoprotein profile.

Adapted from references (166,167). ApoA-I, apolipoprotein A-I; apoB, apolipoprotein B; CE, cholesteryl esters; PL, phospholipids; TG, triglycerides.