Interactions between hepatic lipase andapolipoprotein E gene polymorphisms affectserum lipid profiles of healthy Canadian adults

Kevin C.M. Wood, Morgan D. Fullerton, Ahmed El-Sohemy, and Marica Bakovic

Abstract: The purpose of this study was to assess the individual and interactive effects between hepatic lipase (LIPC;C-514T, G-250A) and apolipoprotein E (APOE) (E2, E3, E4) gene polymorphisms on levels of plasma lipoprotein choles-

terol and triglyceride among healthy, young, Canadian adults (n = 440). All subjects with at least one APOE2 allele hadsignificantly lower low-density lipoprotein cholesterol, total cholesterol, and total cholesterol – high-density lipoproteincholesterol ratio when compared with those with the APOE3 or APOE4 allele. There were significant differences in theLIPC allele and genotype frequencies between Caucasian (n = 207) and Asian (n = 211) individuals, but ethnicity did notcontribute to the variations in circulating lipids. In addition, the lowest triglyceride levels (0.87 ± 0.27 mmol�mL–1) werefound in all APOE2 individuals carrying LIPC-514-CC and LIPC-250-GG genotypes, whereas the highest triglyceride lev-els (1.29 ± 0.34 –1.32 ± 0.32 mmol�mL–1) were found in APOE2 individuals carrying the opposite genotypes, LIPC-514TTand LIPC-250AA. These observations, distinct from the anti-atherogenic effects of APOE2 through the lowering of low-density lipoprotein cholesterol and LIPC on high-density lipoprotein cholesterol, suggest that there is an interactive effectbetween APOE and LIPC genotypes on plasma triglyceride levels. These results provide the basis for further studies on es-tablishing which genotype combinations would be the most protective against hypertriglyceridemia.

Key words: hepatic lipase, apolipoprotein E, polymorphism, SNIPs, lipid profiles.

Resume : Cette etude se propose d’analyser chez 440 jeunes Canadiens en bonne sante les effets individuels et interactifsdes diverses formes du gene de la lipase hepatique (LIPC) (C-514T, G-250A) et de celui de l’apoproteine E (APOE2, E3,E4) sur les concentrations plasmatiques de triglycerides et de cholesterol associees aux lipoproteines. Comparativementaux individus presentant des alleles APOE3 ou APOE4, tous les sujets presentant au moins un allele APOE2 ont une plusfaible concentration de LDL-cholesterol, moins de cholesterol total et un plus faible ratio cholesterol total – HDL-choleste-rol. On observe des differences significatives entre les frequences des alleles LIPC et des genotypes des Caucasiens (n =207) et des Asiatiques (n = 211), mais l’ethnicite n’explique pas les variations des concentrations de lipides dans la circu-lation. En outre, on observe les plus faibles taux de triglycerides (0,87 ± 0,27 mmol�mL–1) chez tous les sujets portantl’allele APOE2 dans les genotypes LIPC-514-CC et LIPC-250-GG et les plus forts taux (1,29 ± 0,34 –1,32 ±0,32 mmol�mL–1) chez les sujets portant l’allele APOE2 mais identifies aux genotypes opposes, LIPC-514TT et LIPC-250AA. Ces observations qui se demarquent des effets anti-atherogenes de APOE2 et de LIPC sur la diminution respectivedes concentrations de LDL-cholesterol et de HDL-cholesterol suggerent une interaction des genotypes APOE et LIPCquant aux taux plasmatiques de triglycerides. Ces resultats constituent la base pour de nouvelles etudes sur les meilleurescombinaisons de genotypes a des fins de lutte contre l’hypertriglyceridemie.

Mots-cles : lipase hepatique, apoproteine E, polymorphisme, polymorphismes mononucleotiques (snips), profil lipidique.

[Traduit par la Redaction]

Introduction

Cardiovascular disease (CVD) remains the leading causeof mortality in industrialized societies such as Canada andthe United States (Klett and Patel 2003). CVD is a multifac-torial disease that is modulated by interactions between en-

vironmental risk factors and multiple predisposing genes(Andreassi et al. 2003). Among these are hepatic lipase(LIPC) and apolipoprotein (APOE). APOE codes for a majorprotein constituent of triglyceride (TG)-rich chylomicronsand very low-density lipoprotein (VLDL) remnants(Hagberg et al. 2000; Pedro-Botet et al. 2001), which medi-ates their binding and uptake via the hepatic low-densitylipoprotein (LDL) receptor, hence mediating clearance fromplasma (Erkkila et al. 2001; Tan et al. 2003; Campos et al.2001; Moreno et al. 2004).

Genetic variation at the APOE locus is an important de-terminant of serum LDL cholesterol (LDL-C) concentration(Corella et al. 2001) and accounts for 5% to 15% of the var-iance in serum total cholesterol (TC) (Nicklas et al. 2002).Three common alleles of the APOE gene have been de-scribed: E2, E3, and E4, which produce 3 isoforms of the

Received 30 October 2007. Accepted 29 April 2008. Publishedon the NRC Research Press Web site at apnm.nrc.ca on 6 June2008.

K.C. Wood, M.D. Fullerton, and M. Bakovic.1 Department ofHuman Health and Nutritional Sciences, University of Guelph,Guelph, ON, N1G 2W1.A. El-Sohemy. Department of Nutritional Sciences, Universityof Toronto, Toronto, ON, M5S 3E2.

1Corresponding author (e-mail: mbakovic@uoguelph.ca).

761

Appl. Physiol. Nutr. Metab. 33: 761–768 (2008) doi:10.1139/H08-054 # 2008 NRC Canada

protein — E2, E3, and E4 (Addya et al. 1997). These iso-forms differ in amino acid sequence at positions 112 and158 (Eichner et al. 2002). E3 has a cysteine at position112 and an arginine at position 158; E2 has a cysteineat both positions; and E4 has an arginine at both posi-tions. APOE allele frequencies vary, with the major allelein all populations being E3 (allele frequency of 0.49–0.91), followed by E4 (allele frequency 0.06–0.37), andE2 (allele frequency 0–0.15) (Campos et al. 2001). Popu-lation studies have shown that plasma TC and LDL-Cconcentrations are lowest in subjects with the E2 allele,intermediate in those with the E3 allele, and highest inthose with the E4 allele (Erkkila et al. 2001; Elosua etal. 2004).

Hepatic lipase (HL), a lipolytic enzyme synthesized pri-marily in hepatocytes, plays a critical role in lipoproteinmetabolism (van’t Hooft et al. 2000; Connelly 1999; Galanet al. 2000; Campos et al. 1995). HL attaches to the sinus-

oidal endothelial surface after being secreted from hepato-cytes and hydrolyzes TGs and phospholipids in plasmalipoproteins (Connelly 1999; Galan et al. 2000; Deeb andPeng 2000). It is thought that increased HL activity is linkedto the generation of small LDLs and is associated with lowhigh-density lipoprotein (HDL) levels owing to its ability toincrease the hepatic uptake of HDL lipids (Grundy et al.1999). There have been several polymorphisms identified inthe LIPC gene, including some causing a rare HL deficiency(Hegele et al. 1992, 1993; Su et al. 2002). Recent studieshave also shown that there are associations between LIPCgene promoter polymorphisms and HL activity (Pihlajamakiet al. 2000; Couture et al. 2000; Guerra et al. 1997; Zambonet al. 1998). A polymorphism in the LIPC gene, whichcauses a C?T substitution at nucleotide 514 (C-514T) is fa-vorable. This allele is associated with decreased HL activity,therefore increasing HDL concentrations, which areinversely associated with the risk of developing CVD

Table 1. General demographic, biochemical, and genotypic characteristics.

Characteristic All participants Men Women pn 440 150 290Age (y) 22.52±2.34 22.79±2.44 22.38±2.28 0.087Weight (kg) 63.39±12.39 71.81±11.00 59.03±10.72 <0.0001Height (cm) 167.5±9.04 175.8±7.10 163.2±6.60 <0.0001BMI (kg�m–2) 22.49±3.31 23.19±2.90 22.13±3.46 0.001Cholesterol (mmol�L–1)

TC 4.19±0.79 4.00±0.74 4.29±0.80 0.0003LDL-C 2.18±0.66 2.19±0.62 2.17±0.68 0.826HDL-C 1.56±0.38 1.35±0.29 1.66±0.38 <0.0001

TG (mmol�L–1) 1.02±0.45 1.05±0.53 1.00±0.45 0.293TC–HDL-C ratio 2.82±0.77 3.08±0.85 2.68±0.68 <0.0001APOE (n (%))

E2 54 (12.3) 15 (10) 39 (13.4) n.s.a

E3 288 (65.4) 107 (71.3) 181 (62.4)E4 98 (22.3) 28 (18.7) 70 (24.2)

E2 allele frequency 0.063 0.050 0.071 n.s.a

E3 allele frequency 0.814 0.850 0.795E4 allele frequency 0.123 0.100 0.134LIPC-C-514T (n (%))

CC 238 (53.4) 87 (58.0) 151 (52.9) n.s.a

CT 168 (38.2) 53 (35.3) 115 (38.6)TT 34 (7.7) 10 (6.7) 24 (8.5)

C allele frequency 0.732 0.757 0.719 n.s.a

T allele frequency 0.268 0.243 0.281LIPC-G-250A (n (%))

GG 235 (53.4) 84 (56.0) 151 (52.1) n.s.a

GA 167 (38.0) 57 (38.0) 110 (37.9)AA 38 (8.6) 9 (6.0) 29 (10.0)

G allele frequency 0.724 0.750 0.710 n.s.a

A allele frequency 0.276 0.250 0.290LIPC haplotype estimates

G-C 0.7203 0.7423 0.7086 n.s.a

G-T 0.0035 0.0068 0.0017A-C 0.0115 0.0134 0.0104A-T 0.2647 0.2366 0.2793

Note: Results are listed as mean ± SD. APOE genotypes are as follows: E2, 2/3, 2/2 and 2/4; E3, 3/3;E4, 3/4 and 4/4. Significant difference (p < 0.05) between men and women were generated by Student’s ttest. n.s., not significant.

aValues determined by Fisher’s exact t test

762 Appl. Physiol. Nutr. Metab. Vol. 33, 2008

# 2008 NRC Canada

(Pihlajamaki et al. 2000; Couture et al. 2000; De Oliveira ESilva et al. 1999). Another polymorphism in the LIPC gene,which causes a G?A substitution at nucleotide 250 (G-250A),is favorably associated with lower HL activity and buoyantLDL particles (Zambon et al. 1998).

The aim of this cross-sectional study was to investigatewhether genotypic differences in the APOE and LIPC genesamong young Canadian adults were associated withdifferences in levels of plasma lipoprotein lipids, LDL-C,HDL-C, TC, and TG.

Materials and methods

Study subjectsThis study consisted of 150 male and 290 female subjects

between the ages of 20 and 29. The exclusion criteria in-cluded pregnant and breast-feeding women, as their metabo-lism and food intake are altered during these conditions. Allsubjects gave written informed consent. This study was ap-

proved by ethics committees from both the University ofGuelph and the University of Toronto.

Clinical characteristicsAfter consenting to participate in the study, subjects were

required to fast for 12 h prior to blood collection. From thisblood sample TC, LDL-C, HDL-C, TC–HDL ratio, and TGwere measured. Height and mass were also measured, andbody mass index (BMI) was calculated. Blood lipids weremeasured using standard clinical procedures at MDS Labo-ratories (Toronto, Ont.). LDL-C was calculated using theFriedewald equation and values were excluded for individu-als with TG concentrations > 4.52 mmol�L–1.

LIPC and APOE genotypingDNA was extracted using QIAamp DNA blood Midi Kits

(Qiagen) following the manufacturer’s protocol. C-514T andG-250A variants of LIPC were genotyped using PCR fol-lowed by cleavage with NlaIII and DraI, respectively, as de-

Table 2. Specific ethnic demographic, biochemical and genotypic characteristics.

Characteristic Asian Caucasian pn 211 207 —Age (y) 22.32±2.26 22.69±2.41 0.103Weight (kg) 59.39±12.12 66.64±10.61 <0.0001Height (cm) 164.7±8.42 170.2±8.76 <0.0001BMI (kg�m–2) 21.80±3.00 22.94±3.10 <0.0002Cholesterol (mmol�L–1)

TC 4.18±0.75 4.16±0.84 0.785LDL-C 2.18±0.65 3.15±0.68 0.552HDL-C 1.56±0.39 1.55±0.37 0.804

TG (mmol�L–1) 0.99±0.42 1.02±0.45 0.452TC–HDL-C ratio 2.82±0.81 2.79±0.70 0.130APOE (n (%))

E2 29 (13.7) 23 (11.1) n.s.a

E3 133 (63.0) 145 (70.1)E4 49 (23.3) 39 (18.8)

E2 allele frequency 0.073 0.056 n.s.a

E3 allele frequency 0.796 0.840E4 allele frequency 0.130 0.100LIPC-C-514T (n (%))

CC 90 (42.7) 138 (66.7) 0.002a

CT 101 (47.9) 60 (29.0)TT 20 (9.4) 9 (4.3)

C allele frequency 0.666 0.812 <0.0001a

T allele frequency 0.334 0.188LIPC-G-250A (n (%))

GG 91 (43.1) 134 (64.7) 0.005a

GA 96 (45.5) 64 (30.9)AA 24 (11.4) 9 (4.3)

G allele frequency 0.659 0.802 <0.0001a

A allele frequency 0.341 0.198LIPC haplotype estimates

H1 (G-C) 0.6540 0.8020 0.038a

H2 (G-T) 0.0048 0.0000H3 (A-C) 0.0119 0.0098H4 (A-T) 0.3293 0.1882

Note: Results are listed as mean ± SD. APOE genotypes are as follows: E2, 2/3, 2/2 and2/4; E3, 3/3; E4, 3/4 and 4/4. Significant difference (p < 0.05) between Asian and Cauca-sian subpopulations were generated by Student’s t test. n.s., not significant.

aValues determined by Fisher’s exact t test.

Wood et al. 763

# 2008 NRC Canada

scribed by Zhiguang et al. (2002). Electrophoresis was per-formed on an 8% polyacrylamide gel, then stained with SybrGreen for 20 min and visualized by UV illumination.

Following PCR, APOE variants were identified by cleav-age with HhaI (Hixson and Vernier 1990). Electrophoresiswas performed on an 18% polyacrylamide gel, and gelswere stained and visualized as above.

Statistical analysesStatistical analyses were carried out using the SPSS statis-

tical software package version 15.0 for the PC, as well asGraphPad Prism 4. To evaluate the effect of the APOEgenotype, subjects were categorized into 3 groups: E2 (E2/E2, E2/E3, and E2/E4), E3 (E3/E3) and E4 (E3/E4 and E4/E4). Differences in age, BMI, mass, height, and plasma lip-ids between sex (male vs. female) and ethnicity (Asian vs.Caucasian) were assessed by Student’s t test. Differences inplasma lipids between ethnic groups, sex, and genotypeswere assessed by one-way analysis of variance (ANOVA),with Tukey’s post hoc test for multiple comparisons. Multi-ple ANOVA (using a general linear model) were performedto determine the presence of an interactive effect of sex andethnicity on APOE, as well as the two LIPC genes onplasma determinants, where significance was computedfrom type III sums of squares with Tukey’s post hoc test toadjust for multiple comparisons. Deviations of genotype fre-quencies from Hardy–Weinberg equilibrium were assessedusing �2 analyses, where 3 and 1 degree of freedom was ap-plied for APOE and LIPC, respectively. Haplotype estimateswere established using the eExpectation-maximization algo-rithm with Hapstat software (available from www.bios.unc.edu/~lin/hapstat/). Significant differences between haplotypeestimates and allele and genotype frequencies were deter-mined using Fisher’s exact t test.

Results

Characteristics of the study populationTable 1 provides a summary of the demographic,

genotypic, and biochemical characteristics of all 440participants in the study. The mean age of the participantswas 22.52 years. The mean values of plasma TC, LDL-C,HDL-C, and TG were 4.19, 2.18, 1.56, and 1.02 mmol�L–1,respectively. The frequencies of the LIPC and APOE allelesare similar to the relative frequencies reported in the litera-

ture and all polymorphisms are in Hardy–Weinberg equili-brium (Eichner et al. 2002; Couture et al. 2000; de Andradeet al. 2004). The mean ages of the men and women were22.79 and 22.38 years, respectively. Mass and height weresignificantly lower in women than in men (p < 0.0001) andBMI was significantly higher in men than in women (p =0.001). The levels of HDL-C were significantly higher inwomen than in men and TC and TC–HDL ratio were signif-icantly lower in women than in men (p < 0.001).

Table 2 details the subject and biochemical characteristicsseparately for Asian (n = 211) and Caucasian (n = 207) in-dividuals, the two predominant ethnic backgrounds in thisstudy. Although the mean weight, height and BMI were sig-nificantly different between ethnic groups, the plasma lipiddeterminants were unaltered. Interestingly, we determinedthere to be no difference in APOE allele or genotype fre-quency between Asian and Caucasian subjects, however, theallele and genotype frequencies for both LIPC G-250A andC-514T polymorphisms were significantly varied betweenethnic groups. In spite of an altered LIPC genotype distribu-tion between the two ethnic groups (Table 1), further analy-ses revealed no difference in any of the plasma parametersfor either LIPC polymorphism (results not shown).

Analysis by ethnicity and sexDifferences in plasma lipids according to the APOE geno-

type for all subjects are shown in Table 3. The E2 genotypehad significantly lower TC (p < 0.0001), LDL-C (p <0.0001) and TC/HDL ratio (p < 0.001) than either the E3 orE4 groups, with no significant difference existing betweenthe E3 and E4 groups. When plasma lipids were analyzedbased on APOE genotype within the different ethnic back-grounds, a slightly altered profile emerged (Table 4). In theAsian subpopulation, there were no differences in TC; how-ever, the E2 group had significantly lower LDL-C comparedwith E3 and E4 (p values were 0.02 and 0.03, respectively),whereas the E4 group had a significantly altered TC–HDLratio (p = 0.03). Interestingly, the Caucasian subpopulationwith the E2 genotype had significantly lower TC, LDL-C,and TC–HDL ratio (p < 0.001, p < 0.0001, and p < 0.03,respectively) compared with the E3 and E4 groups. Compar-isons between the two ethnic groups also revealed that TCand LDL-C are affected in the E2 individuals, whereas lipidparameters for the E3 and E4 genotypes remain unaltered.

A summary of the plasma lipid levels according to the in-

Table 3. Plasma TG and cholesterol levels of all participants according to APOE genotype.

Genotype

Variable E2 E3 E4 p

n 54 288 98 —Age (y) 22.63±2.65 22.42±2.27 22.74±2.39 0.472BMI(kg�m–2) 22.52±3.50 22.41±3.05 22.73±3.91 0.701Cholesterol (mmol�L–1)

TC 3.80±0.66 4.21±0.79 4.33±0.79 <0.0001LDL-C 1.75±0.61 2.21±0.64 2.32±0.64 <0.0001HDL-C 1.60±0.37 1.56±0.38 1.53±0.38 0.526

TG (mmol�L–1) 0.99±0.32 1.07±0.48 1.02±0.45 0.421TC–HDL-C ratio 2.48±0.66 2.83±0.77 2.97±0.76 0.001

Note: Results are listed as mean ± SD. Significant difference (p < 0.05) were determined by AN-OVA and are relative to the E2 genotype.

764 Appl. Physiol. Nutr. Metab. Vol. 33, 2008

# 2008 NRC Canada

teraction of APOE genotype and sex is provided in Table 5.These results demonstrate significant differences betweenthe E2, E3, and E4 groups and differences according to sex(p < 0.0001). There was a significant difference in LDL-C(p < 0.0001) according to the APOE genotype and a signifi-cant difference in HDL-C (p < 0.0001) according to sex. Wenext examined the effect of APOE genotype and ethnicityon plasma lipid determinants; however, there was no effectof ethnicity, nor any interactive effect of ethnicity and

genotype determined for any of the parameters (data notshown). Two-way ANOVA also revealed no significantalterations among any of the plasma lipids for either theC-514T or G-250A LIPC genotypes when examining par-ticipants according to genotype-by-ethnicity or genotype-by-sex interactions (results not shown).

Interactive effect of hepatic lipase and ApoE genotypesWe next performed analyses to distinguish combined ef-

Table 4. Plasma TG and cholesterol levels of ethnic groups according to APOE genotype.

Genotype

Variable E2 E3 E4

AsianLDL-C 1.87±0.67 2.21±0.64 (0.03) 2.29±0.60 (0.02)HDL-C 1.68±0.42 1.55±0.37 1.52±0.41TG, mmol�L–1 0.96±0.28 0.98±0.42 1.04±0.50TC–HDL-C ratio 2.49±0.69 2.84±0.82 2.96±0.80 (0.03)

CaucasianTC 3.52±0.50a 4.20±0.83 (0.001) 4.42±0.86 (0.0001)LDL-C 1.55±0.45£ 2.18±0.65 (0.0001) 2.36±0.73 (0.00001)HDL-C 1.52±0.29 1.56±0.37 1.54±0.36TG, mmol�L–1 0.99±0.33 0.99±0.49 1.14±0.48TC–HDL-C ratio 2.40±0.60 2.79±0.68 (0.03) 3.00±0.76 (0.003)

Note: Results are listed as mean ± SD. Analyses by ANOVA, where p values shown in parentheses representdifferences relative to E2 group.

aSignificant difference (p < 0.05) between E2 Asian and E2 Caucasian individuals.

Table 5. APOE genotype-by-sex interaction on plasma lipids.

Genotype

E2 E3 E4 p value

Variable Men Women Men Women Men Women Genotype Sex G�S

TC (mmol�L–1) 3.64±0.63 3.87±0.67 4.05±0.75 4.30±0.80 3.99±0.71 4.47±0.77 <0.0001 <0.0001 0.457LDL-C (mmol�L–1) 1.78±0.63 1.75±0.60 2.22±0.61 2.20±0.66 2.26±0.58 2.34±0.67 <0.0001 0.899 0.784HDL-C (mmol�L–1) 1.44±0.32 1.66±0.38 1.37±0.28 1.67±0.39 1.24±0.27 1.64±0.36 0.343 <0.0001 0.393TG (mmol�L–1) 0.95±0.28 1.00±0.34 1.05±0.54 0.97±0.40 1.07±0.57 1.07±0.45 0.367 0.176 0.792TC–HDL ratio 2.69±0.83 2.40±0.57 3.08±0.84 2.68±0.69 3.33±0.86 2.83±0.67 0.001 <0.0001 0.772

Note: Results are listed as mean ± SD. Analyses by multiple ANOVA, where significant difference was set at the p value of <0.05 for genotype and sex,both individually and for the interactive effect.

Table 6. The effects of LIPC genotypes on plasma TG within APOE genotypes.

APOE LIPC-C-514T n TG LIPC-G-250A n TG

E2 CC 29 0.87±0.27a GG 29 0.87±0.27b

CT 19 1.06±0.30 GA 18 1.06±0.31TT 6 1.32±0.34 AA 7 1.29±0.32

E3 CC 154 1.01±0.49 GG 152 1.00±0.48CT 115 1.02±0.44 GA 115 1.03±0.46TT 19 0.90±0.20 AA 21 0.90±0.20

E4 CC 55 1.15±0.57 GG 54 1.14±0.57CT 34 0.99±0.33 GA 34 1.01±0.33TT 9 0.89±0.27 AA 10 0.86±0.26

p value APOE � LIPC 0.029c 0.032c

Note: Results are listed as mean ± SD.aSignificance at p < 0.003 compared with C-514T genotype TT, generated by one-way ANOVA.bSignificance at p < 0.004 compared with G-250A genotype AA, generated by one-way ANOVA.cSignificant interaction between APOE and LIPC genotype, where p values for interactive effect of genotype were

generated by multiple ANOVA.

Wood et al. 765

# 2008 NRC Canada

fects of APOE and LIPC genotypes on plasma lipids. No in-teractive effects for TC, LDL-C, HDL-C, or TC–HDL ratiowith either sex vs. genotype or ethnicity vs. genotype wereestablished (results not shown). On the other hand, all indi-viduals with E2 genotype (independently from ethnicity andsex) demonstrated significant variations in TG levels be-tween LIPC-514 CC vs. TT genotypes (p = 0.003), as wellas LIPC-250 GG vs. AA genotypes (p = 0.004) (Table 6).The E2 � 514-CC and E2 � 250-GG individuals had lowerplasma TG (0.87 ± 0.27 mmol�L–1) compared with E2 �514-TT and E2 � 250-AA individuals (1.32 ± 0.34 and1.29 ± 0.32 mmol�L–1, respectively). TG levels did not sig-nificantly vary with LIPC genotype within the E3 or E4 sub-groups. It is also noticeable that LIPC heterozygous (514-CTand 250-GA) E2 and E4 individuals have plasma TG levelsintermediate to LIPC homozygotes, demonstrating a cleargene–dosage effect. In addition, the effect of APOEgenotypes on plasma TG levels within each LIPC subgroup(Table 7) demonstrated that LIPC-514-TT and LIPC-250-AAindividuals that also have at least 1 E2 allele, and have higherplasma TG (1.29–1.32 mmol�L–1) when compared with theindividuals that carry E3 or E4 alleles (0.86–0.89 mmol�L–1).Finally, comparing the overall gene–gene interaction on TGlevels via multiple ANOVA, there was a significant differ-ence in TG levels among the APOExLIPC(C-514T) geno-types (p = 0.029) and the APOExLIPC(G-250A) genotypes(p = 0.032) (Table 6).

Discussion

The purpose of this study was to investigate the interactionbetween APOE and LIPC gene polymorphisms in modifyinglipoprotein cholesterol and TG content in young, healthy,Canadian adults. The main finding is that having at least 1APOE2 allele and the LIPC(C-514T) and LIPC(G-250A)polymorphisms are significant determinants of total plasmaTG levels. The optimal gene combination for lower plasmaTG appears to be APOE2 and LIPC-514-CC or APOE2 andLIPC-250-GG.

The effects of APOE on serum cholesterol content wereconsistent with previous studies (Hixson 1991; Ortega et al.2005; Srinivasan et al. 1993), with higher LDL-C and TCpresent in subjects with the E3 or E4 allele compared withsubjects with the E2 allele. There was an interactive effectbetween APOE genotypes and sex and generally more ele-

vated HDL-C in females than in males. These results are inagreement with the Framingham Study (Wilson et al. 1994;Lahoz et al. 2001), suggesting that women possessing an E2allele may be among the least susceptible to CVD. In addi-tion, we demonstrate that the TC and LDL-C levels in Cau-casian subjects with the E2 genotype are lower than inAsian subjects with the same genotype, indicating that eth-nicity may play a role in modulating the beneficial effectsof the E2 genotype.

When analysis was performed on the LIPC gene, no sig-nificant effects of sex and ethnicity on plasma LDL-C, TC,and HDL-C were established. The allelic variations at theLIPC locus, together with APOA and APOC loci, however,are known determinants of total HDL-C and could accountfor as much as 25% in its variability (Dugi et al. 2001;Lambert et al. 2001; Dugi et al. 2000). Over-expression ofHL decreases HDL-C, whereas HL deficiency results in theopposite (Santamarina-Fojo et al. 1998). For the C-514Tpolymorphism, the T allele is associated with decreased en-zyme activity and increased HDL-C (Hegele et al. 1993;Dugi et al. 2001), and for the G-250A polymorphism, the Aallele also leads to decreased activity and higher HDL-C. Inthis study, the LIPC polymorphisms did not have any signif-icant effect on HDL-C levels, which is in agreement with arecent study that also describes a young, normolipidemicpopulation (Jimenez-Gomez et al. 2008). This study, how-ever, clearly established that the interactions between LIPCand APOE gene loci to influence total plasma TG levels,for which a plausible mechanism could be proposed basedon their common effect on uptake and clearance of TG-richlipoprotein particles, in addition to their separate effects onHDL-C (by LIPC) and LDL-C (by APOE) (Santamarina-Fojo et al. 2004; Al-Haideri et al. 1997).

Generally, HL is considered a complex protein that lowersplasma levels of both the proatherogenic LDL-C and theantiatherogenic HDL-C, so that its net effect on plasma lipo-proteins is not always obvious (Santamarina-Fojo et al.2004). HL also has a ligand-binding, non-catalytic functionin improving lipoprotein interactions by cell surface recep-tors and proteoglycans that reduces atherogenic risk(Santamarina-Fojo et al. 2004). It is also present in the arte-rial wall, where these ligand-type interactions could be athe-rogenic (Santamarina-Fojo et al. 2004). In peripheral tissues,and the arterial wall, apoE is also involved in the uptake ofTG-rich particles via cell-surface proteoglycans that are non

Table 7. The effects of APOE genotypes on the variations in plasma TG within LIPC genotypes.

LIPC-C-514T APOE n TG LIPC-G-250A APOE n TG

CC E2 29 0.87±0.27a GG E2 29 0.87±0.27a

E3 154 0.99±0.47 E3 152 0.99±0.45E4 55 1.14±0.57 E4 54 1.14±0.57

CT E2 19 1.06±0.30 GA E2 18 1.06±0.31E3 115 1.01±0.44 E3 115 1.03±0.46E4 34 0.97±0.33 E4 34 1.01±0.31

TT E2 6 1.32±0.34b AA E2 7 1.29±0.32b

E3 19 0.89±0.20 E3 21 0.90±0.20E4 9 0.89±0.29 E4 10 0.86±0.26

Note: Results are listed as mean ± SD.aSignificance difference (p < 0.03) between E2 vs. E4 genotypes, generated by one-way ANOVA.bSignificance difference (p < 0.003) between E2 vs. E3 and E2 vs. E4, generated by one-way ANOVA.

766 Appl. Physiol. Nutr. Metab. Vol. 33, 2008

# 2008 NRC Canada

receptor-mediated (Al-Haideri et al. 1997). It is not clearhow these separate roles of HL and apoE are related withtheir polymorphisms, but two independent investigationsdemonstrated that APOE 4/3 vs. APOE 3/3 in young men(Bergeron and Havel 1996) and type 2 diabetic individualswith LIPC-514-TT (Baum et al. 2005) have an impairedclearing of intestinal (chylomicron) and hepatic (VLDL)TG-rich lipoprotein remnants. Also, in obese individuals, anE4 allele increases the risk of elevated plasma TG (Gueguenet al. 1989). Type 2 diabetics also have an unfavorable ef-fect of E2 and E4 alleles on fasting TG (Reznik et al.2002), and the risk for diabetic nephropathy between type 2diabetics is increased in individuals with LIPC-514-TT incombination with E2 or E4 (Baum et al. 2005). In our study,we established that in young, healthy, adult Canadians theAPOE–LIPC genotypes are associated with TG metabolism,and that the optimal genotype combination for lower TG lev-els is APOE2 (lower LDL-C and TC) with LIPC-514CC and(or) LIPC-250GG (both increase HL activity), therefore sug-gesting that the most interactive effects of these gene poly-morphisms was likely at the level of more efficient clearanceof TG-rich lipoproteins (Jimenez-Gomez et al. 2008).

In conclusion, our results corroborate findings from otherpopulation studies that the APOE2 allele also has a signifi-cant cholesterol-lowering effect among normolipidemicyoung Canadian adults. We demonstrate that APOE2’scholesterol-lowering effect may be influenced by ethnicityand sex, and that the frequencies of the 2 LIPC polymor-phisms varied significantly between the Asian and Caucasiansub-populations, without any effect on the lipid parameters in-vestigated. We have also provided new evidence that theAPOE and LIPC gene loci may have an interactive, modifyingeffect on total plasma TG levels in young Canadians, whichprovides a foundation for further studies to identify protectivegenotype combinations against hypertriglyceridemia.

AcknowledgmentsThis study was supported by funds from the Advanced

Foods and Materials Network, part of the National Centresfor Excellence.

ReferencesAddya, K., Wang, Y.L., and Leonard, D.G.B. 1997. Optimization

of apolipoprotein E genotyping. Mol. Diagn. 2: 271–276.PMID:10462619.

Al-Haideri, M., Goldberg, I.J., Galeano, N.F., Gleeson, A., Vogel,T., Gorecki, M., et al. 1997. Heparan sulfate proteoglycan-mediated uptake of apolipoprotein E-triglyceride-rich lipoproteinparticles: a major pathway at physiological particle concentra-tions. Biochemistry, 36: 12766–12772. doi:10.1021/bi9631024.PMID:9335533.

Andreassi, M.G., Botto, N., Cocci, F., Battaglia, D., Antonioli, E.,Masetti, S., et al. 2003. Methylenetetrahydrofolate reductasegene C677T polymorphism, homocysteine, vitamin B12 andDNA damage in coronary artery disease. Hum. Genet. 112:171–177. PMID:12522558.

Baum, L., Ng, M.C.Y., So, W.-Y., Lam, V.K.L., Wang, Y., Poon,E., et al. 2005. Effect of hepatic lipase –514CT polymorphismand its interactions with apolipoprotein C3 –482CT and apolipo-protein E exon 4 polymorphisms on the risk of nephropathy in

Chinese type 2 diabetic patients. Diabetes Care, 28: 1704–1709.doi:10.2337/diacare.28.7.1704. PMID:15983323.

Bergeron, N., and Havel, R.J. 1996. Prolonged postprandial re-sponses of lipids and apolipoproteins in triglyceride-rich lipo-proteins of individuals expressing an apolipoprotein epsilon 4allele. J. Clin. Invest. 97: 65–72. doi:10.1172/JCI118408.PMID:8550852.

Campos, H., D’Agostino, M., and Ordovas, J.M. 2001. Gene-diet in-teractions and plasma lipoproteins: role of apolipoprotein E and ha-bitual saturated fat intake. Genet. Epidemiol. 20: 117–128. doi:10.1002/1098-2272(200101)20:1<117::AID-GEPI10>3.0.CO;2-C.PMID:11119301.

Campos, H., Dreon, D.M., and Krauss, R.M. 1995. Associations ofhepatic and lipoprotein lipase activities with changes in dietarycomposition and low density lipoproteins subclasses. J. LipidRes. 36: 462–472. PMID:7775858.

Connelly, P.W. 1999. The role of hepatic lipase in lipoprotein me-tabolism. Clin. Chim. Acta, 286: 243–255. doi:10.1016/S0009-8981(99)00105-9. PMID:10511296.

Corella, D., Tucker, K., Lahoz, C., Coltell, O., Cupples, L.A.,Wilson, P., et al. 2001. Alcohol drinking determines the effect ofthe APOE locus on LDL-cholesterol concentrations in men: theFramingham Offspring Study. Am. J. Clin. Nutr. 73: 736–745.PMID:11273848.

Couture, P., Otvos, J.D., Cupples, L.A., Lahoz, C., Wilson, P.W.,Schaefer, E.J., et al. 2000. Association of the C-514T poly-morphism in the hepatic lipase gene with variations in lipo-protein subclass profiles: the Framingham Offspring Study.Arterioscler. Thromb. Vasc. Biol. 20: 815–822. PMID:10712408.

de Andrade, F.M., Silveira, F.R., Arsand, M., Antunes, A.L.S.,Torres, M.R., Zago, A.J., et al. 2004. Association between –250G/A polymorphism of the hepatic lipase gene promoter andcoronary artery disease and HDL-C levels in a southern Brazi-lian population. Clin. Genet. 65: 390–395. doi:10.1111/j.0009-9163.2004.00243.x. PMID:15099346.

Deeb, S.S., and Peng, R. 2000. The C-514T polymorphism in thehuman hepatic lipase gene promoter diminishes its activity. J.Lipid Res. 41: 155–158. PMID:10627514.

De Oliveira E Silva, E., Kong, M., Han, Z., Starr, C., Kass, E.M.,et al. 1999. Metabolic and genetic determinants of HDL metabo-lism and hepatic lipase activity in normolipidemic females. J.Lipid Res. 40: 1211–1221. PMID:10393206.

Dugi, K.A., Amar, M.J.A., Haudenschild, C.C., Shamburek, R.D.,Bensadoun, A., Hoyt, R.F., Jr., et al. 2000. In vivo evidence forboth lipolytic and nonlipolytic function of hepatic lipase in the me-tabolism of HDL. Arterioscler. Thromb. Vasc. Biol. 20: 793–800.PMID:10712405.

Dugi, K.A., Brandauer, K., Schmidt, N., Nau, B., Schneider,J.G., Mentz, S., et al. 2001. Low hepatic lipase activity is anovel risk factor for coronary artery disease. Circulation, 104:3057–3062. doi:10.1161/hc5001.100795. PMID:11748100.

Eichner, J.E., Dunn, S.T., Perveen, G., Thompson, D.M., Stewart,K.E., and Stroehla, B.C. 2002. Apolipoprotein E polymorphismand cardiovascular disease: a HuGE review. Am. J. Epidemiol.155: 487–495. doi:10.1093/aje/155.6.487. PMID:11882522.

Elosua, R., Ordovas, J.M., Cupples, L.A., Fox, C.S., Polak, J.F., Wolf,P.A., et al. 2004. Association of APOE genotype with carotid ather-osclerosis in men and women: the Framingham heart study. J. Li-pid Res. 45: 1868–1875. doi:10.1194/jlr.M400114-JLR200.PMID:15258198.

Erkkila, A.T., Sarkkinen, E.S., Lindi, V., Lehto, S., Laakso, M.,and Uusitupa, M.I. 2001. APOE polymorphism and the hyper-tryglyceridimic effect of dietary sucrose. Am. J. Clin. Nutr. 73:746–752. PMID:11273849.

Wood et al. 767

# 2008 NRC Canada

Galan, X., Robert, M.Q., Llobera, M., and Ramirez, I. 2000. Secre-tion of hepatic lipase by perfused liver and isolated hepatocytes.Lipids, 35: 1017–1026. doi:10.1007/s11745-000-0613-z. PMID:11026623.

Grundy, S.M., Vega, G.L., Otvos, J.D., Rainwater, D.L., and Cohen,J.C. 1999. Hepatic lipase activity influences high density lipo-protein subclass distribution in normotriglyceridaemic men:genetic and pharmacological evidence. J. Lipid Res. 40: 229–234.PMID:9925651.

Gueguen, R., Visvikis, S., Steinmetz, J., Siest, G., and Boerwinkle,E. 1989. An analysis of genotype effects and their interactionsby using the apolipoprotein E polymorphism and longitudinaldata. Am. J. Hum. Genet. 45: 793–802. PMID:2816943.

Guerra, R., Wang, J., Grundy, S.M., and Cohen, J.C. 1997. A hepa-tic lipase (LIPC) allele associated with high plasma concentra-tion of high density lipoprotein cholesterol. Proc. Natl. Acad.Sci. U.S.A. 94: 4532–4537. doi:10.1073/pnas.94.9.4532. PMID:9114024.

Hagberg, J.M., Wilund, K.R., and Ferrell, R.E. 2000. APOE geneand gene–environment effects on plasma lipoprotein-lipid levels.Physiol. Genomics, 4: 101–108. PMID:11120871.

Hegele, R.A., Little, J.A., Vezina, C., Maguire, G.F., Tu, L.,Wolever, T.S., et al. 1993. Hepatic lipase deficiency: clinical,biochemical, and molecular genetic characteristics. Arterioscler.Thromb. 13: 720–728. PMID:8485124.

Hegele, R.A., Tu, L., and Connelly, P.W. 1992. Human hepatic li-pase mutations and polymorphisms. Hum. Mutat. 1: 320–324.doi:10.1002/humu.1380010410. PMID:1301939.

Hixson, J.E. 1991. Apolipoprotein E polymorphisms affect athero-sclerosis in young males. Pathobiological determinants ofatherosclerosis in youth (PDAY) research group. Arterioscler.Thromb. 11: 1237–1244. PMID:1680392.

Hixson, J.E., and Vernier, D.T. 1990. Restriction isotyping of hu-man apolipoprotein E by gene amplification and cleavage withHhaI. J. Lipid Res. 31: 545–548. PMID:2341813.

Jimenez-Gomez, Y., Perez-Jimenez, F., Marın, C., Gomez, P.,Moreno, R., Delgado, J., et al. 2008. The –250/A polymorphismin the hepatic lipase gene promoter influences the postprandiallipemic response in healthy men. Nutr. Metab. Cardiovasc. Dis.18: 173–181. PMID:17399967.

Klett, E.L., and Patel, S. 2003. Genetic diseases against noncholes-terol sterols. Curr. Opin. Lipidol. 14: 341–345. doi:10.1097/00041433-200308000-00001. PMID:12865730.

Lahoz, C., Schaefer, E.J., Cupples, L.A., Wilson, P.W.F., Levy, D.,Osgood, D., et al. 2001. Apolipoprotein E genotype and cardiovas-cular disease in the Framingham heart study. Atherosclerosis, 154:529–537. doi:10.1016/S0021-9150(00)00570-0. PMID:11257253.

Lambert, M.S., Avella, M.A., Berhane, Y., Shervill, E., and Botham,K.M. 2001. The differential hepatic uptake of chylomicron rem-nants of different fatty acid composition is not mediated by hepa-tic lipase. Br. J. Nutr. 85: 575–582. PMID:11348572.

Moreno, J.A., Perez-Jiminez, F., Marin, C., Gomez, P., Perez-Marinez, P., Moreno, R., et al. 2004. Apolipoprotein E genepromoter –219G?T polymorphism increases LDL-cholesterolconcentrations and susceptibility to oxidation in response to adiet rich in saturated fat. Am. J. Clin. Nutr. 80: 1404–1409.PMID:15531693.

Nicklas, B.J., Ferrell, R.E., Bunyard, L.B., Berman, D.M., Dennis,K.E., and Goldberg, A.P. 2002. Effects of apolipoprotein Egenotype on dietary-induced changes in high-density lipoproteincholesterol in obese postmenopausal women. Metabolism, 51:853–858. doi:10.1053/meta.2002.33337. PMID:12077730.

Ortega, H., Castilla, P., Gomez-Coronado, D., Garces, C., Benavente,M., Rodriguez-Artalejo, F., et al. 2005. Influence of apolipopro-tein E genotype on fat-soluble plasma antioxidants in Spanishchildren. Am. J. Clin. Nutr. 81: 624–632. PMID:15755832.

Pedro-Botet, J., Schaefer, E.J., Bakker-Arkema, R.G., Black, D.M.,Stein, E.M., Corella, D., et al. 2001. Apolipoprotein E genotypeaffects plasma lipid response to atorvastatin in a gender specificmanner. Atherosclerosis, 158: 183–193. doi:10.1016/S0021-9150(01)00410-5. PMID:11500190.

Pihlajamaki, J., Karjalainen, L., Karhapaa, P., Vauhkonen, I.,Taskinen, M.R., Deeb, S.S., et al. 2000. G-250A substitution inpromoter of hepatic lipase gene is associated with dyslipidemiaand insulin resistance in healthy control subjects and in mem-bers of families with familial combined hyperlipidemia. Arter-ioscler. Thromb. Vasc. Biol. 20: 1789–1795. PMID:10894818.

Reznik, Y., Morello, R., Pousse, P., Mahoudeau, J., and Fradin, S.2002. The effect of age, body mass index, and fasting triglycer-ide level on postprandial lipemia is dependent on apolipoproteinE polymorphism in subjects with non-insulin-dependent diabetesmellitus. Metabolism, 51: 1088–1092. doi:10.1053/meta.2002.34696. PMID:12200750.

Santamarina-Fojo, S., Gonzalez-Navarro, H., Freeman, L., Wagner,E., and Nong, Z. 2004. Hepatic lipase, lipoprotein metabolism, andatherogenesis. Arterioscler. Thromb. Vasc. Biol. 24: 1750–1754.doi:10.1161/01.ATV.0000140818.00570.2d. PMID:15284087.

Santamarina-Fojo, S., Haudenschild, C., and Amar, M. 1998. Therole of hepatic lipase in lipoprotein metabolism and athero-sclerosis. Curr. Opin. Lipidol. 9(3): 211–219. PMID:9645503.

Srinivasan, S.R., Ehnhom, C., Wattigney, W., and Berenson, G.S.1993. Apolipoprotein E polymorphism and its association withserum lipoprotein concentrations in black versus white children:the Bogalusa heart study. Metabolism, 42: 381–386. doi:10.1016/0026-0495(93)90091-2. PMID:8487659.

Su, Z., Zhang, S., Nebert, D.W., Zhang, L., Huang, D., Hou, Y., etal. 2002. A novel allele in the promoter of the hepatic lipase isassociated with increased concentration of HDL-C and de-creased promoter activity. J. Lipid Res. 43: 1595–1601. doi:10.1194/jlr.M200046-JLR200. PMID:12364543.

Tan, C.E., Tai, E.S., Tan, C.S., Chia, K.S., Lee, J., Chew, S.K., et al.2003. APOE polymorphism and lipid profile in three ethnic groupsin the Singapore population. Atherosclerosis, 170: 253–260.doi:10.1016/S0021-9150(03)00232-6. PMID:14612205.

Van’t Hooft, F.M., Lundahl, B., Ragogna, F., Karpe, F., Olivecrona,G., and Hamsten, A. 2000. Functional characterization of 4 poly-morphisms in promoter region of hepatic lipase gene. Arterios-cler. Thromb. Vasc. Biol. 20: 1335–1339. PMID:10807751.

Wilson, P.W., Myers, R.H., Larson, M.G., Ordovas, J.M., Wolf,P.A., and Schaefer, E.J. 1994. Apolipoprotein E alleles, dyslipi-demia, and coronary heart disease. The Framingham Offspringstudy. JAMA, 272: 1666–1671. doi:10.1001/jama.272.21.1666.PMID:7966894.

Zambon, A., Deeb, S.S., Hokanson, J.E., Brown, B.G., andBrunzell, J.D. 1998. Common variants in the promoter of the he-patic lipase gene are associated with lower levels of hepatic li-pase activity, buoyant LDL, and higher HDL2 cholesterol.Arterioscler. Thromb. Vasc. Biol. 18: 1723–1729. PMID:9812910.

Zhiguang, S., Zhang, S., Nebert, D.W., Zhang, L., Huang, D., Hou,Y. et al. 2002. A novel allele in the promoter of the hepatic li-pase is associated with increased concentration of HDL-C anddecreased promoter activity. J. Lipid Res. 43: 1595–1601.

768 Appl. Physiol. Nutr. Metab. Vol. 33, 2008

# 2008 NRC Canada