In the present study, we tested the effects of MGLL gene polymorphisms on plasma LDL-C and LDL particle size following an n-3 PUFA supplementation. The MGLL gene has been previously shown to be differentially expressed in peripheral blood mononuclear cells (PBMCs) between subjects responders and non-responders defined on the basis of the plasma TG levels variation in pre- versus post- n-3 PUFA supplementation. 18 SNPs were genotyped to cover 100% of the common genetic variations. To our knowledge, this is the first study to examine the influence of MGLL gene variations on LDL-C levels and LDL particle size in relation to an n-3 PUFA supplementation.
Near half of the subjects increased their LDL-C levels (55%) and the plasma LDL-C variation ranged from -1.7 mmol/L to +1.0 mmol/L, which shows the large inter-individual variability of the LDL-C response to the n-3 PUFA supplementation observed in this cohort. In response to dietary interventions, many studies also found an important inter-individual variability of the plasma LDL-C. In studies with the National Cholesterol Education Panel step 2 diets, Schaefer et al. found a plasma LDL-C variation ranging between +13% and -55% between participants . A review also showed that LDL-C varied from +17% to +46% in hypertriglyceridemic subjects receiving 4 g/day of n-3 PUFA . We also found that the LDL particle size was not statistically different in pre- versus post-supplementation period in our study, while many studies found an increase in the LDL particle size following an n-3 PUFA supplementation, as described earlier. It has been described earlier that the LDL particle size is inversely related to plasma TG concentrations . In a study by Griffin et al., altering the n-6/n-3 ratio by giving PUFA enriched diet to subjects has been found to decrease TG levels as well as the proportion of small, dense LDL particles . In addition, Kelley et al. showed a decrease in TG levels and a increase in LDL particle size following a DHA supplementation in hypertriglyceridemic men . The TG levels in our cohort decreased by 11%, as reported previously , compared to a mean of 25% decreased following 3–4 g/day n-PUFA in normolipemic subjects in a meta-analysis . This may potentially explain the lack of difference in LDL particle size after the n-3 PUFA supplementation. Dietary intakes in relation to LDL particle size variation have been studied by Faghihnia et al. reporting that a low-fat, high-carbohydrate diet reduces LDL particle size . However, in the present study, results remained unchanged after further adjustment for dietary intakes. The lack of difference in effect of the n-3 PUFA supplementation on LDL particle size may also be attributable to the large inter-variability observed in the plasma TG levels in our study cohort, as previously reported .
We verified the independent effects of supplementation and genotype as well as the supplementation*genotype interaction effects on LDL-C and LDL particle size. No supplementation or genotype effects were observed for LDL-C. However, six SNPs (rs782440, rs6776142, rs555183, rs782444, rs6787155, rs1466571) of MGLL showed significant supplementation*genotype effects on LDL-C, suggesting that these variants may modulate the LDL-C response to an n-3 PUFA supplementation. Four SNPs of MGLL (rs782440, rs13076593, rs549662, rs541855) were associated with LDL particle size. The T/T carriers for the rs782440 SNP, the C/C for rs13076593, the A/G + G/G genotypes combined for rs549662 and the C/C genotype for rs541855 had a beta value of 0 or higher for LDL particle size comparatively to the other genotype groups. Three SNPs showed a supplementation*genotype interaction (rs782440, rs9877819, rs13076543). The rs782440 SNP is present in all these statistical associations. To our knowledge, genetic variants of MGLL have only been studied once in the literature in relation to metabolic outcomes in humans. An association study of Harismendy et al. showed that in a population of obese individuals as well as non-obese controls, three intervals of rare variants in the MGLL gene sequence are associated with BMI (promoter, intron 2 and intron 3) .
The MGLL enzyme is known to hydrolyse 2-AG, which is one of the central components of the endocannabinoid signaling network . The activation of the cannabinoid receptors by 2-AG have an impact on energy homeostasis by stimulating appetite, promoting lipid storage and reducing energy expenditure. They have also been associated with metabolic changes associated with obesity and metabolic syndrome . Circulating 2-AG levels in the plasma have been correlated positively with BMI and waist girth, as well as with plasma TG levels , while another study found that 2-AG levels were increased by 52% in obese women . These higher levels of 2-AG in obese individuals may result of a reduced enzymatic degradation by the MGLL enzyme. Although we did not find any associations between MGLL SNPs and BMI or waist circumference, these measures have been strongly associated to plasma lipid levels  and LDL particle features . However, a study with MGLL-KO mice showed that despite an elevation of 2-AG levels, animals did not have enhanced lipid storage, increased appetite or decreased energy expenditure .
The allele frequency disparities observed between the different genotype groups and the positive responder/negative responder status for LDL-C variation suggests that individuals with different genotypes for these SNPs may respond differently to an n-3 PUFA supplementation.